There was a time, 30 or 40 years ago, when construction litigation was a byword for expense and delay, and where the costs were often out of all proportion to the sums at stake. Subsequently, thanks in part to compulsory construction adjudication, which has reduced the number of construction cases that go through to a final trial, and the careful case management by TCC judges of those cases which do, construction litigation has become a much more efficient and cost-effective method of dispute resolution. But occasionally, circumstances conspire to create a construction case with echoes of the bad old days. Unfortunately, this is one such case.

The appellant is an American-qualified architect, who was a friend and former neighbour of the respondents. Gratuitously, she provided assistance to the respondents when they wanted to undertake major landscaping works (including structural elements) in their North London garden (“the garden project”). There was a falling-out which led the respondents to commence proceedings against the appellant in the TCC for breach of contract and/or negligence. The appellant made a Part 36 offer in the sum of £25,000 three weeks after the start of proceedings, which was not accepted.

Following the trial of Preliminary Issues, the existence of any contract was rejected. Although it was found that the appellant owed the respondents a duty of care (a finding subsequently upheld by the Court of Appeal), this court made plain that, in these particular circumstances, the duty of care related to only such professional services as the appellant in fact provided; in other words, she could have no liability in respect of any alleged omissions. That meant that what the appellant actually did was critical to any claim in negligence against her. Following numerous further interlocutory skirmishes and a 5-day trial, the judge concluded that the appellant had in fact provided very few services and had not been negligent in providing any of them. The claim failed in its entirety.

The appellant’s costs were presented to the judge in the eye-watering amount of £724,265 (and even that was incomplete, because it excluded some items such as the costs of the earlier appeal). She sought assessment on an indemnity basis. The respondents argued, and the judge agreed, that costs should be assessed on the standard basis. The appellants’ appeal against that decision on costs, together with the respondents’ notice, raised three distinct issues:

Whether this was a case in which the respondents’ pursuit of what were said to be “speculative, weak, opportunistic or thin claims” could properly be described as out of the norm such as to warrant an order for indemnity costs.

Whether the respondents’ failures to accept and subsequently to beat the appellant’s Part 36 offer, made at a very early stage in the proceedings, also meant (either separately or taken cumulatively with the pursuit of these particular claims) that an order for indemnity costs was warranted.

The relevance, if any, of the fact that the appellant’s approved costs budget was said to be £415,000, but that any assessment on the indemnity basis would start at the appellant’s actual costs figure of not less than £724, 265.

The factual background to this case has already been set out in detail on three separate occasions: by Mr Alexander Nissen QC (sitting as a Deputy High Court Judge) in his ruling on the Preliminary Issue ([2016] EWHC 40 (TCC)); by the Court of Appeal in their judgment upholding his decision but modifying the duty of care to reflect what the appellant actually did, rather than what it was alleged she omitted to do ([2017] EWCA Civ 254); and in the substantive judgment of Mr Martin Bowdery QC (sitting as a

In 2013, the respondents were considering the garden project and obtained a quotation of £155,837, plus a planting budget of £19,785 (both exclusive of VAT) from Mark Enright of the Landscape Garden Company Limited. The respondents did not immediately proceed with this quotation because of their concerns about the costs. Following some casual conversations with the appellant, she became involved in the garden project between March and July 2013. In April 2013 (contrary to the evidence of the first respondent) Mr Nissen QC subsequently found that a budget figure of £130,000 was discussed between the appellant and the first respondent.

Over the next few weeks, a considerable amount of piling and other groundworks was carried out at the site. By early July the first respondent had become concerned about the budget. At a meeting on 8 July, he said that he had understood that the budget was £78,000, not £130,000. He denied that he had ever agreed the £130,000. In consequence of this, on 9 July 2013, the appellant terminated her relationship with the respondents. At no stage did she make any claim for fees in connection with the garden project.

The respondents did not comply with the TCC pre-action protocol, but commenced proceedings, alleging breaches of contract and a duty of care at common law, on 5 March 2015. The claim was for approximately £300,000, most of which was a global claim (explained below). On 26 March 2015, the appellant made an offer in accordance with CPR Part 36 to settle the case for £25,000. This offer was not accepted either within 21 days, or at all. Instead, on 31 March 2015, the respondents made a counteroffer in the sum of £220,000.

The claim was unusual because of the nature of the relationship between the parties, and the fact that such services as were provided by the appellant were provided free of charge. It appears that, at the first CMC, Edwards-Stuart J was sceptical about the legal basis of the claims, and he ordered Preliminary Issues to address the existence of the alleged contract and the alleged duty of care. As noted above, Mr Nissen QC decided those Preliminary Issues in a judgment dated 15 January 2016. Although he found the existence of a duty of care in tort, the respondents lost their case that there was a contract, and also suffered a major reverse on the arguments about the budget advice. Thereafter, on 5 February 2016, the respondents made a Part 36 offer in the sum of £150,000, which was £70,000 less than their offer the previous March.

On 16 March 2016, the appellant was granted permission to appeal, and the TCC proceedings were stayed. The Court of Appeal judgment, referred to above, was dated

Thereafter, there were an extraordinary number of further interlocutory applications and hearings. Most, but not all, of these were generated by the respondents. In particular, it took them two attempts (September and December 2017) to amend their Particulars of Claim in the light of the Court of Appeal judgment. In important respects, their new case was diametrically opposed to their original case: for example, having originally denied that they were given a budget figure as high as £130,000, the respondents were now obliged to argue that this figure was much too low.

In its final form, the shape of the pleaded claim was as follows. Section F(2) of the Amended Particulars of Claim was concerned with alleged failures in respect of inspection and supervision. These allegations were made by reference to a Scott Schedule. It was axiomatic that these defects, and other alleged “non-conformances” with the Enright design, were (or ought to have been) apparent to the appellant by 9 July 2013 (which was when she ceased her relationship with the respondents). Section F(3) set out particulars of negligence in relation to the few design drawings provided by the appellant. Section F(4) comprised allegations of failure in respect of budgeting and cost control.

The Particulars of Loss at paragraph 33 set out the respondents’ primary claim, which was for every penny that the respondents had spent over and above the Enright quotation. This has been referred to subsequently as ‘the global claim’. As amended, it amounted to £172,224.53. The global claim had always been part of the original Particulars of Claim (albeit in a different sum) but, when giving permission for the amendments to it, O’Farrell J was unimpressed enough to remark that the global claim “should come with a government health warning”. Paragraph 33(2) of the Particulars of Loss set out the secondary claim for alleged overpayments made in excess of what was said to have been the proper value of the works and materials.

The claims were denied in comprehensive detail in the re-amended defence. Although the pleading is lengthy, it comprised the following essential components:

The appellant denied that she had undertaken any detailed design work, and therefore denied that there had been (or could have been) any negligent design.

The appellant denied that she had undertaken any periodic inspections, but further averred that, in any event, none of the alleged defects were in existence (or could have been seen) at the time that she ceased her involvement in the garden project on 9 July 2013.

The appellant denied the claim for negligence in relation to the budget; again the principal point being that no budgeting services were undertaken at all.

The appellant denied the global claim both as a matter of principle (it was averred that the claim was fundamentally flawed because it paid no regard to which elements of the overall sum claimed could be said have been caused by and/or were the foreseeable consequence of the breach alleged), and on the facts.

As the case proceeded towards trial, an attempt was made to address the issue of costs budgeting. O’Farrell J made some orders in relation to costs budgeting on 1 December 2017. Although the respondents argued on appeal that the consequence of these was that the appellant’s approved costs budget was £415,000 excluding VAT, there was no order to that effect. Only some elements of the anticipated costs were the subject of the order; other important elements (such as the anticipated costs of experts) were not. Paragraph 11 of O’Farrell J’s order of 1 December 2017 demonstrated that the incurred costs (which even at that stage ran to hundreds of thousands of pounds) were not agreed by the parties. She also ordered that revised costs budgets were to be provided. It does not appear that this ever happened.

By the time of the trial, each side had two experts (an architect and a quantity surveyor (Footnote: ¹)). However, the reports were not served until 6 June 2018, only a few weeks before the trial in July 2018. That was far too late to allow a sensible appraisal of the reports prior to trial. Moreover, on 12 April 2018, Fraser J had ordered the appointment of a joint expert to address the question of whether or not the defects and nonconformances were or should have been apparent to the appellant on or before 9 July 2013. It appears that the joint report, prepared by Mr Christopher Milnes, was not provided until 2 July 2018, just a fortnight before the start of the trial. Mr Milnes concluded from the documents that he was shown (which largely comprised the photographs which were in the possession or control of the respondents) that all but one of the items in the respondents’ schedule were not apparent or in existence before or on 9 July 2013. Despite that, these and all the other claims noted above were pursued at the trial.

Between April 2017 (when the Court of Appeal modified the order of Mr Nissen QC) and the start of the trial on 16 July 2018, the respondents made two further offers:

On 11 January 2018, they sent a Part 36 offer offering to settle the claim for £45,000 (one fifth of the value of their original offer made before the decisions of Mr Nissen

On 22 March 2018 they sent a Calderbank offer offering to settle for £48,000 and 60% of their costs. Because of the huge costs which had been racked up by then on both sides, this offer meant that, even if the appellant had accepted it, the respondents would have been significantly out of pocket.

I note below the important parts of the judge’s main judgment.

In his general observations, the judge criticised the respondents’ “scatter gun approach” which he described as “unhelpful” [24]. In addition, at [27] – [30] the judge set out in detail how and why he concluded that the first respondent’s evidence was “unsatisfactory”, “argumentative and often inconsistent with the contemporary documentation”, and “very difficult to understand”. By contrast, at [35] – [37] the judge recorded his view that the appellant “answered questions clearly and concisely”, how “her evidence particularly compared with Peter Burgess’ evidence, was impressive and largely consistent with the contemporaneous documentation”, and that, “on all factual issues, save one, I prefer her evidence to the evidence of the Burgesses”.

The judge reminded himself at [18] – [19] and again at [37(xiii)] that, following the decision of the Court of Appeal, what mattered was “what services the [appellant] actually provided”, in order that he could “provide a definitive statement of the nature and extent of the duties owed by [the appellant].”

As to what the appellant had done by way of design, the judge said that the appellant’s drawings were never intended to be detailed design drawings or used for construction purposes: [37(xvii)]. He went on at [37(xviii)] to reject the allegations relating to design changes, concluding that “on any view, these alleged changes to the design of the garden were not produced negligently”.

The judge was equally damning about the allegation that the appellant had undertaken periodic inspections. He said:

Having identified what limited services the appellant had actually provided, the judge then went on to address the specific allegations of breach of duty by reference to the pleadings, starting at [38]. He rejected the allegations in respect of design at [40] – [55], saying that the claim for negligent design and project management “lacks credibility and conviction” [44], noting that prior to the hearing in the Court of Appeal, the respondents had accepted that the drawings themselves were not produced negligently and that the original design case had turned on the allegation that further drawings should have been produced. When that case was no longer open to the respondents as a result of the Court of Appeal judgment, the respondents then reneged on their previous position and suggested that the appellant’s drawings had been defective, after all. The judge rejected that changed case in its entirety.

The judge rejected the specific allegations in respect of inspection at [56] – [86]. At [56] he said that it was reasonable that the appellant did not carry out any inspections of the ongoing structural and groundworks at the site and that no other inspections were carried out negligently. He also noted at [57(i)] that many of the alleged defects which it was said that the appellant should have seen had not even been supported by the respondents’ own architectural expert (Mr Armes). He considered that the respondents had wrongly assumed that any item of bad workmanship on the part of the contractor would automatically be reflected in a claim against the appellant for negligent inspection. Of course, this claim also suffered from the fundamental problem, identified by the joint surveying expert, that all but one of these defects were not apparent on or before 9 July 2013.

As to the related allegations that the appellant had failed to ensure conformance with the Enright design, the judge was even more forthright, saying at [88] that “these claims are, on any view, hopeless and no claim in negligence against the [appellant] in respect of these claims should have been pleaded, let alone pursued”. He went on in the same paragraph to describe this as a “thoroughly unmeritorious claim”.

In relation to the allegations in respect of the budget, at [90] – [99] the judge rejected the allegations both in relation to the budget itself and in relation to the interim payments made to Hardcore. He noted at [92] that “as with their design allegations following the preliminary issue judgment the [respondents] advanced a new and wholly inconsistent case that the budget of £130,000 was always unachievable and this project could not have been completed for less than £188,000.”

From [100] onwards, the judge addressed the global claim. He described this claim at [103] as having “many weaknesses, leaving aside my findings that the [appellant] did not act in breach of any duty owed to the [respondents]”. He went on to note at [104] that the global cost “includes many items which the [appellant] cannot, on any view, be liable for”. He went on to say that “to claim that the [appellant] is liable for this global claim offends common sense and I find it wholly unsupported by the evidence which I have heard and read.”

Accordingly, all of the respondents’ claims against the appellant failed, many for more than one reason. It followed that the respondents were prima facie liable to the appellant for her costs of the action. The issue was whether those costs should be assessed on an indemnity or a standard basis.

In an ex tempore judgment handed down on 26 February 2019, the judge concluded that the costs should be assessed on the standard basis. When he considered the claim for indemnity costs by reference to CPR 44.2 and the relevant authorities, he rejected the contention that the respondents’ pre-action conduct and their non-compliance with the pre-action protocol justified an award for indemnity costs: see [12] – [16]. I note that no point now arises in relation to those two elements of the judge’s costs judgment.

As to the conduct of the litigation, dealt with in the costs judgment from [17] – [22], the judge addressed various specific matters such as the confused nature of the pleadings, the making of allegations without expert evidence, the shambolic nature of the disclosure, and the “haphazard and spray gun manner” of the case on defects. Those

As to the unmeritorious nature of the claims, at [20] the judge said in relation to the global claim that it was “not a claim that was hopeless from the beginning; it was a claim which had to be considered at trial and dealt with at trial.” Similarly, at [23], the judge said that, although he had been critical in the main judgment of elements of the respondents’ case, he considered that the case had been won at trial and was not a

The judge dealt with the appellant’s Part 36 offer from [25] onwards. He noted that, in contrast with the position of a claimant who makes a Part 36 offer and then subsequently beats it, a defendant in the same position was not automatically entitled under the CPR to indemnity costs. At [26] he correctly noted that, as part of his general discretion, the fact that the appellant had beaten her own offer was an important matter which he should take into account in considering whether or not the appellant was entitled to indemnity costs. But he did not appear to consider that issue further because, in the very next paragraph at [27], the judge concluded that this was a case in which costs should be assessed on the standard basis.

CPR 36.17 provides as follows:

Subject to rule 36.21, this rule applies where upon judgment being entered—

a claimant fails to obtain a judgment more advantageous than a defendant’s Part 36 offer; or

judgment against the defendant is at least as advantageous to the claimant as the proposals contained in a claimant’s Part 36 offer. (Rule 36.21 makes provision for the costs consequences following judgment in certain personal injury claims where the claim no longer proceeds under the RTA or EL/PL Protocol.)

For the purposes of paragraph (1), in relation to any money claim or money element of a claim, “more advantageous” means better in money terms by any amount, however small, and “at least as advantageous” shall be construed accordingly.

Subject to paragraphs (7) and (8), where paragraph (1)(a) applies, the court must, unless it considers it unjust to do so, order that the defendant is entitled to—

costs (including any recoverable pre-action costs) from the date on which the relevant period expired; and

interest on those costs.”

In contrast to r.36.17(3), a claimant who beats his or her Part 36 offer is entitled, pursuant to r.36.17(4) to a wide range of benefits, including an order for indemnity costs, unless the court considers it unjust to make such an order.

CPR 44.2 sets out the court’s discretion as to costs:

The court has discretion as to –

whether costs are payable by one party to another; (b) the amount of those costs; and (c) when they are to be paid.

If the court decides to make an order about costs –

the general rule is that the unsuccessful party will be ordered to pay the costs of the successful party; but

the court may make a different order…

whether a party has succeeded on part of its case, even if that party has not been wholly successful; and

any admissible offer to settle made by a party which is drawn to the court’s attention, and which is not an offer to which costs consequences under Part 36 apply.

conduct before, as well as during, the proceedings and in particular the extent to which the parties followed the Practice Direction – Pre-Action Conduct or any relevant pre-action protocol;

whether it was reasonable for a party to raise, pursue or contest a particular allegation or issue;

the manner in which a party has pursued or defended its case or a particular allegation or issue; and

whether a claimant who has succeeded in the claim, in whole or in part, exaggerated its claim.” 36. The basis of assessment is set out in CPR 44.3:

on the standard basis; or

on the indemnity basis, but the court will not in either case allow costs which have been unreasonably incurred or are unreasonable in amount. (Rule 44.5 sets out how the court decides the amount of costs payable under a contract.)

only allow costs which are proportionate to the matters in issue. Costs which are disproportionate in amount may be disallowed or reduced even if they were reasonably or necessarily incurred; and

resolve any doubt which it may have as to whether costs were reasonably and proportionately incurred or were reasonable and proportionate in amount in favour of the paying party. (Factors which the court may take into account are set out in rule 44.4.)

Where the amount of costs is to be assessed on the indemnity basis, the court will resolve any doubt which it may have as to whether costs were reasonably incurred or were reasonable in amount in favour of the receiving party.”

The general approach to applications for indemnity costs, made by a successful defendant who has beaten his or her own offer, can be discerned from three cases: Reid Minty (A Firm) v Taylor [2001] EWCA Civ 1723; Kiam II v MGN (No 2)[2002] EWCA Civ 66;and Excelsior Commercial and Industrial Holdings Limited v Salisbury Hammer Aspden and Johnson (A Firm) [2002] EWCA Civ 879.

In Reid Minty, the successful defendant sought indemnity costs relying, in part, on a letter it had written to the claimant inviting the claimant to abandon the case. May LJ said:

There will be many cases in which, although the defendant asserts a strong case throughout and eventually wins, the court will not regard the claimant's conduct of the litigation as unreasonable and will not be persuaded to award the defendant indemnity costs. There may be others where the conduct of a losing claimant will be regarded in all the circumstances as meriting an order in favour of the defendant of indemnity costs. Offers to settle and their terms will be relevant and, if they come within Part 36, may, subject to the court's discretion, be determinative.”

In Kiam v MGN, those paragraphs were considered by Simon Brown LJ. He said:

Both judgments were considered in Excelsior. As to the passages in Kiam, the Lord Chief Justice said:

I take those two examples only for the purpose of illustrating the fact that there is an infinite variety of situations which can come before the courts and which justify the making of an indemnity order. It is because of that that I do not respond to Mr Davidson's submission that this court should give assistance to lower courts as to the circumstances where indemnity orders should be made and circumstances when they should not. In my judgment it is dangerous for the court to try and add to the requirements of the CPR which are not spelt out in the relevant parts of the CPR. This court can do no more than draw attention to the width of the discretion of the trial judge and re-emphasise the point that has already been made that, before an indemnity order can be made, there must be some conduct or some circumstance which takes the case out of the norm. That is the critical requirement.

In this case I am satisfied there was such a circumstance as I have indicated. That being so, it seems to me that the appeal in respect of the indemnity issues should be rejected. In view of what I have already said about the other issue, in my judgment this appeal should be dismissed.” 41.In the same case, Waller LJ said of the passages in Kiam:

Although these judgments are 20 years old, and there have been a number of changes to the underlying rules (including significant changes to Part 36) since, the applicability of the guidance provided in Excelsior has recently been reaffirmed by this court in Shalaby v London North West Health Care NHS Trust [2018] 3 Costs LR 585. In addition, there have been many cases where, in the circumstances, a claimant’s unreasonable failure to accept offers of settlement have warranted an order for costs on the indemnity basis, such as Franks v Sinclair (Costs) [2006] EWHC 3656 (Ch); Southwark LBC v IBM UK Ltd (Costs) [2011] EWHC 653 (TCC); Barr v Biffa Waste Services Ltd (costs) [2011] EWHC 1107 (TCC); and Optical Express Ltd and Others v Associated Newspapers Limited [2017] EWHC 2707 (QB).

In short, therefore, taking the CPR and these authorities together, the position is that, in contrast to the position of a claimant, a defendant (such as the appellant in the present case) who beats his or her own Part 36 offer, is not automatically entitled to indemnity costs. But a defendant can seek an order for indemnity costs if he or she can show that, in all the circumstances of the case, the claimant’s refusal to accept that offer was unreasonable such as to be “out of the norm”. Moreover, if the claimant’s refusal to accept the offer comes against the background of a speculative, weak, opportunistic or thin claim, then an order for indemnity costs may very well be made. That is what happened in Excelsior.

There is a separate strand of authority concerned with speculative, weak, opportunistic or thin claims. It has long been the position that a defendant’s eventual defeat of such claims can give rise to an order for indemnity costs. In Three Rivers District Council v The Governor and Company of the Bank of England[2006] EWHC 816 (Comm), at paragraph 25, Tomlinson J (as he then was) summarised the position:

There are a number of cases where costs have been awarded on an indemnity basis because of the weakness of the claimant’s underlying claims: see by way of example Wates Construction Limited v HGP Greentree Alchurch Evans Limited [2006] BLR 45. In my summary of these principles in Elvanite Full Circle Limited v AMEC Earth and Environmental (UK) Limited [2012] EWHC 1643 (TCC), I referred to Wates as an example of a ‘hopeless’ claim, because on the facts of the case, that is what it was. I did not intend by that shorthand to indicate any sort of gloss on the conventional description of claims which were ‘speculative, weak, opportunistic or thin’ giving rise to the possibility of indemnity costs.

As outlined above there are three distinct issues on this appeal. The first is whether, even leaving aside the question of the Part 36 offer, the judge erred in principle in failing to find that the respondents’ conduct was out of the norm, such that indemnity costs should have been ordered. In my view, the central element of this debate is whether or not, prior to trial, the respondents (or their advisors) should have realised that these were speculative/weak claims which were most unlikely to succeed and that, in pursuing them to trial, their conduct was out of the norm and should have been recognised by an award of indemnity costs in favour of the appellant.

The second issue is whether, when seen against the background of the respondents’ claims, the fact that the appellant bettered her own Part 36 offer meant, in all the circumstances of the case, that the judge again erred in principle in failing to make an award of indemnity costs. Much of the debate, both before the judge and in the written material provided to this court, concerned the CPR and the plain distinction in the treatment of a claimant who beats its own Part 36 offer, on the one hand, and a defendant who beats such an offer, on the other. In my view, this focus was misguided. For good or ill, the CPR plainly makes that distinction. The right question was and remains whether, in all the circumstances of this case, the respondents’ failures to accept and then to beat the offer meant that indemnity costs should have been awarded, not automatically, but in the exercise of the judge’s discretion pursuant to CPR Part 44.3.

Finally, there is the separate issue, raised by way of the Respondents’ Notice, as to whether an award of indemnity costs should not be made because of the gap between what is said to have been her approved cost budget of £415,000 and her actual costs of not less than £724,265.63. That raises potential questions as to the overlap, if any, between the cost budgeting regime on the one hand, and the basis of assessment of costs after trial, on the other.

I deal with each of those issues in turn below. I do so against the background of the test to be applied to appeals concerned with costs, articulated by Sir Murray Stuart-Smith in Roache v Newsgroup Newspapers [1998] EMLR 161, when he said at page 172:

There are therefore only two ways in which this court may interfere with a costs decision. The first is if there has been an error in law. The second, which is generally much harder to establish, is based on the submission that the discretion was exercised in a manner which led to an unjust or perverse result. In the present case, Mr Cohen (who had the burden of this aspect of the appeal) made plain that he relied on the former route, namely the error of law, and not the latter. I address the first two issues on that basis; the third issue did not arise before the judge.

As I have already said, there can be no issue with the judge’s conclusions that the respondents’ conduct both in relation to the pre-action period and in relation to the noncompliance with the pre-action protocol does not amount to conduct which is out of the norm. I take the same view in relation to the specific matters that were raised by the appellant in relation to the conduct of the litigation itself, summarised at paragraph 30 above. These events were all a product of the way in which this litigation was fought on both sides, and the judge’s refusal to apportion particular blame one way or the other for those matters was a matter for his discretion. No error of principle in the judge’s approach is disclosed.

There is, however, much more force in the appellant’s over-arching point about the judge’s failure to address the speculative/weak nature of these claims. As noted above, at [20] and [23] of the costs judgment, the judge seemed to approach the merits issue on the basis that, because there had had to be a trial in order for the lack of merit in these claims to be finally determined, there was no entitlement to indemnity costs. He appeared to consider that an order for indemnity costs was only appropriate where it could be shown with hindsight that costs had been unnecessarily incurred. I do not accept that this was the right approach as a matter of principle. Indemnity costs are, for example, routinely ordered in favour of a vindicated defendant when allegations of fraud are dismissed at trial. Obviously, there are many cases in which the strength of one side’s position, or the flaws in the other side’s case, only become apparent at trial: Bank of Ireland v Watts[2017] EWHC 2472 (TCC), to which the judge referred, was just such a case. But that is a different point: in that case, the claims themselves could not be described as prospectively weak or speculative.

In addition, although the judge was referred in the written submissions to Excelsior and Three Rivers, during oral argument it appears that he was asked to focus primarily on whether the claims could be said in hindsight to be ‘hopeless’, rather than whether they should have been seen by the respondents at any time prior to the trial as speculative or weak. Contrary to Mr Oram’s post-hearing note, I consider that there is a substantive difference between the two formulations.

None of that should be regarded as a criticism of the judge. It is quite clear that the essential focus of Mr Flannery’s submissions was to ask the judge to conclude that, in hindsight, the claims were hopeless because they were all dismissed at trial, and to order indemnity costs in consequence. That may have been because Mr Oram was arguing the converse. But however that focus came about, it was misplaced: the judge should instead have been asked to consider whether, at any time following the commencement of the proceedings, a reasonable claimant would have concluded that the claims were so speculative or weak or thin that they should no longer be pursued. I note that this approach was suggested for the first time in the appellant’s supplemental skeleton argument provided a few days before the appeal, which was doubtless the result of Mr Cohen’s industry and knowledge of the authorities.

It seems to me that, although this point of principle was raised late, it was undoubtedly the right question, and the judge erred in not addressing it. He was also led into error by both counsel’s focus on whether or not the respondents’ claims were hopeless. That too was not the right test. For these reasons, it is necessary for this court to consider the question posed in the preceding paragraph.

In my view, for the reasons set out below, the answer to the question is plain. No later than one month after the handing down of the judgment by the Court of Appeal on 7 April 2017 (i.e. by 7 May 2017) the respondents, having had time to consider the implications of the Court of Appeal judgment, should have realised that the remaining claims were so speculative/weak that they were very likely to fail, and should not be pursued any further.

First, I consider that that was the cumulative effect of the decision of Mr Recorder Nissen QC (no claim in contract, rejection of the original budget advice claim), and then the Court of Appeal judgment itself. Prior to the latter, because of the finding of a duty of care on the part of the appellant, the respondents had an arguable (if not very strong) case both in relation to design and in relation to inspection, that focused not on what the appellant had done, but on what it was said that the appellant had failed to do. The Court of Appeal’s judgment, although in one sense a modest modification of the original finding of a duty, emphasised that the duty related only to the things that the appellant had actually done, not the things which it was suggested she should have done but omitted to do.

The critical impact of this change should have been only too obvious to the respondents themselves, because (as the residents) they will have known almost as well as the appellant herself how little she had in fact done. Furthermore, the detrimental effect of this change on the respondents’ case can also be tracked in two ways as the events unfolded thereafter: the amendments to the respondents’ pleaded case, and their realisation of the weakness of their own claims.

Thus, in their pleaded case on design, the respondents had originally asserted that, although there was nothing wrong with the drawings that the appellant produced, she should have produced other, more detailed drawings. Once that argument (based as it was on omissions) was no longer open to them, the respondents had to create a new case which suggested that her drawings were inadequate, after all. That case plainly arose out of necessity (i.e. it was all that the respondents could say, in the light of the Court of Appeal’s judgment), rather than an objective and reasoned view of the merits (i.e. that the allegations were objectively justified).

The speculative/weak nature of that design case was demonstrated for all to see at the trial: see paragraph 23 above. But the failure of a case that was diametrically opposed to that which the respondents had originally argued cannot have come as a surprise to anyone. Moreover, it was no answer for Mr Oram to say that some parts of the design case were supported by expert evidence: such expert evidence as there was did not support the claim as pleaded, did not primarily come from an architect, and was served much too late to be the basis of the amended allegations. The principal reason why the new and wholly inconsistent design claim failed was not because of the judge’s rejection of the limited, late expert evidence which supported it, but because, once the omissions claim had gone, he found that the new claim “lacked credibility and conviction”. That was in effect a finding that the new design claim had been, since its formulation in 2017, speculative/weak.

There was a similar volte face in relation to the budget figure of £130,000 and the claims arising from it. As the judge noted in his judgment at [92], the judgment on the Preliminary Issues, as confirmed by the Court of Appeal, required the respondents to advance “a new and wholly inconsistent case” that the budget figure of £130,000, far from being too much (which was why the parties had fallen out in the first place), was actually too little. Again, that change of case was driven by necessity rather than the merits of the underlying allegation.

Again, it can therefore have been no surprise that the claims arising out of the budget (including the global claim) failed at trial. As noted at paragraphs 26-27 above, the judge said that he could not understand how the budget claim “could be seriously maintained” [91], and that the global claim which derived from it “offends against common sense” [104]. The judge had therefore found these claims had necessarily always been speculative/weak.

The inspection claim, by reference to the Scott Schedule, was also significantly impacted by the Court of Appeal’s judgment, because the appellant had done so little on site. In my view, that claim always was (and should have been appreciated by the respondents to be) speculative/weak. The respondents, as the residents of the property, would have known better than anyone that (as the joint expert eventually confirmed by reference to the photographs that the respondents belatedly disclosed) all but one of these defects were not apparent on site on or prior to 9 July 2013, when the appellant’s involvement ceased. Secondly, as the judge found at [57(ii)], the claim was in any event based on the false premise that any item of bad workmanship automatically gave rise to an allegation of negligent supervision against the appellant. The related claim for non-conformance with the Enright design was regarded by the judge at [88] of the costs judgment as being “hopeless” and a claim which he said should not have been pleaded, let alone pursued (see paragraphs 24-25 above). That finding was more than enough to trigger indemnity costs, at least in relation to these parts of the claim.

In short, contrary to the new points put forward in Mr Oram’s post-hearing note, I am bound to conclude that the claims maintained and/or modified after the Court of Appeal judgment did not need a long and expensive trial for it to have been apparent to the respondents and their advisors that they were, at the very least, speculative/weak claims.

I am also confident that, following the Court of Appeal judgment, the respondents were themselves aware that their claims now faced significant difficulties. Although this was not dealt with at all by the judge, this can be seen in their own Part 36 offers. Having sought £220,000 before the decision in the Court of Appeal, thereafter they were prepared to accept one fifth of that (£45,000) and, subsequently, they were prepared to take even less when they offered to accept £48,000 and just 60% of their costs. Those significant reductions in the amount that the respondents were prepared to take is, on the facts of this case, eloquent testimony to the fact that the respondents must have known that, after the Court of Appeal judgment, their claims were speculative/weak and that all that really mattered now was costs. From then on, the respondents found themselves in an absurd position, where they were incurring hundreds of thousands of pounds in costs, solely in order to try and recover some of those costs from the appellant. The respondents should have called a halt, because their underlying claims were speculative/weak, but they failed to do so.

I have asked myself why, in all those circumstances, these speculative/weak claims were pursued to trial. The answer may very well lie in the judge’s comment at [108] of his main judgment: that the decision to continue was borne out of the respondents’ desire “to punish the appellant for her alleged negligent mistakes rather than seek fair and reasonable compensation for her alleged mistakes”. An irrational desire for punishment unlinked to the merits of the claims themselves is precisely the sort of conduct which the court is likely to conclude is out of the norm.

In summary therefore, although the judge dealt with the specific criticisms that were made of various aspects of the respondent’s conduct of the litigation, he was at no time invited to stand back and, on the evidence before him, consider whether or not there came a time when the respondents knew or ought to have known that their claims were speculative/weak and therefore likely to fail. That was the relevant question. For the reasons that I have given, I consider that, if he had asked himself the question, the judge would have concluded that such a time came not later than one month after the Court of Appeal judgment (namely by 7^th May 2017). It was, in my view, out of the norm for these respondents to continue to pursue the appellant with these speculative/weak claims, each of which the judge himself described in very similar terms, beyond that date.

Accordingly, if my Lady and my Lord agree, I would allow the appeal in relation to the conduct of the respondents. For the reasons that I have given, I consider that the pursuit of these claims from 7 May 2017 onwards was out of the norm such as to justify an order for indemnity costs.

Although I have concluded that the pursuit of these speculative and weak claims justified an award of indemnity costs from 7 May 2017 onwards, the respondents’ rejection of the appellant’s Part 36 offer remains relevant for two separate reasons.

First, even if my analysis in Section 6, was wrong and the respondents’ pursuit of these claims was not on its own sufficient to justify an award of indemnity costs, their refusal of the appellant’s Part 36 offer was a separate matter which, when seen against the background of these particular claims, might justify an order for indemnity costs (as per Excelsior). Secondly, even if I was right about the respondents’ conduct, the respondents’ failure to beat the offer might be relevant to the basis of the assessment of costs between the Part 36 offer and 7 May 2017, the date noted above.

As I have already indicated, prior to the appeal hearing itself, there was a good deal of debate about the absence of an automatic entitlement on the part of a defendant to indemnity costs in circumstances such as these. Mr Flannery went so far as to say in his original skeleton for the purposes of the appeal that this was “scandalous” and, on any view, “an oversight” on the part of the Rules Committee. In my view, the making of these colourful submissions ended up blinding both Mr Flannery and the judge from the real position.

Those submissions about the Rules were misconceived. Part 36 has always been designed to provide that a claimant who beats his or her offer has an automatic entitlement to indemnity costs (unless that can be shown to be unjust) whilst a defendant has no such automatic right. Although some commentators consider this misalignment to be unjustified, it is right to note that Part 36 has been the subject of detailed consideration and amendment on at least two occasions since it was first drafted, and at no stage has it been thought appropriate to adjust this misalignment. Indeed, when the issue was the subject of specific consultation in 2006, there were mixed views as to whether or not the position of a defendant should be brought into line with that of a claimant (Footnote: ²) and no changes were made.

Thus, as things presently stand, the CPR is clear. There is no automatic entitlement on the part of a defendant to indemnity costs if that defendant beats its own Part 36 offer.

In the course of his oral submissions, Mr Cohen put this argument in a more nuanced fashion. He said that, following the more recent changes to the CPR, and in particular the renewed emphasis on proportionality, this court should conclude that there was now a presumption in favour of a defendant who beats his or her own Part 36 offer that they should be entitled to indemnity costs, or that in some way, a defendant’s beating of its own offer should be given “a pre-eminent status” in any claim for indemnity costs.

I agree that the changes to the CPR in respect of proportionality are important. Their effect has been neatly summarised by Marcus Smith J in Bohinc v Malmsten [2019] EWHC 1386 (Ch) as follows:

In other words, the proportionality of the overall bill claimed determined the rigour of the subsequent assessment of costs. If the costs were proportionate, then the costs judge would simply need to consider whether each individual item was reasonably incurred at a reasonable cost. If the costs were disproportionate, on the other hand, then a higher (necessity) standard was used. Only if that cost item was necessary would the reasonable costs of that item be allowed. If the incurring of a cost was necessary, and the amount itself reasonable, there was no further ability to reduce the overall costs bill by reference to proportionality or, indeed, any other measure. The final bill could remain disproportionate to the matter at issue.

The present rules are very different. It is quite clear, from the express wording of CPR 44.3(2)(a) that there may be a reduction in costs on grounds of disproportionality even if those costs were reasonably or necessarily incurred.”

I accept, therefore, that a claimant in the position of these respondents, unless there is an order against them for indemnity costs, will be entitled to take all sorts of points as to the proportionality of the costs incurred by the appellant which would not have been available before the rule changes. But I think it goes much too far to say that, in some way, this would disincentivise a claimant from accepting a Part 36 offer because, as Mr Cohen put it, “it reduces the stick element in failing to accept an offer”. The recent changes might reduce the ultimate amount of the defendant’s costs payable by a losing claimant, but it is hardly likely that the possibility of arguing for such reductions would amount to a cogent reason why, months if not years before those costs were even incurred, a claimant would reject a defendant’s otherwise reasonable Part 36 offer.

More fundamentally, I cannot accept Mr Cohen’s submission that these changes could or should give rise to any sort of presumption in favour of a defendant who beats his or her own Part 36 offer. That seems to me to amount to a backdoor rewrite of the CPR. It would be contrary to Reid Minty,Kiam, and Excelsior. I am quite clear that the changes to the CPR in respect of proportionality do not warrant such a radical course.

However, as the judge correctly noted, the absence of an automatic entitlement is the beginning, rather than the end, of the analysis. The fact that a defendant has beaten his or her own Part 36 offer is plainly a matter of importance in the exercise of the court’s discretion under CPR Part 44. The authorities repeatedly emphasise that. The question is how the judge addressed the Part 36 offer when exercising his discretion in the present case.

The problem as I see it is that, having properly recorded at [26] that the respondents’ failure to beat the appellant’s offer was an important matter in the exercise of his discretion, in the very first sentence of the next paragraph of his costs judgment, the judge said that he did not think that this was a case for indemnity costs. In other words, at no point in his costs judgment did the judge say why this important factor did not, in all the circumstances of the case, lead him to exercise his discretion in favour of an order for indemnity costs or, alternatively, why an award on the standard basis was appropriate, notwithstanding the respondents’ rejection of the appellant’s early offer. It was simply not a matter that the judge addressed.

When a defendant beats its own Part 36 offer, the court should always consider whether, in consequence, the claimant’s conduct in refusing that offer took the case out of the norm. Sometimes it will; sometimes it won’t. Mr Cohen articulated the question that had to be asked in these terms:

Again, on the facts of the present case, I consider that the answer is plain. The appellant had undertaken some work for people that she regarded as her friends, free of charge. That work took a matter of weeks. It involved one or two drawings and some visits to site. Following what the appellant saw as the respondents’ unreasonable (not to say mendacious) attitude towards the £130,000 figure which she had discussed with the first respondent not once but twice, she ceased her involvement in the garden project.

Between then and the commencement of proceedings there was a good deal of acrimony. The appellant must have been all too aware of how difficult any litigation might be. Accordingly, she made an offer at the outset of the proceedings of £25,000. That offer was made just three weeks after the proceedings had started but, if it had been accepted, it would also have carried with it a significant costs liability on her part, in view of the costs both sides had incurred from July 2013 to March 2015.

That offer was made for two reasons. First, it was made in the hope that the respondents would take it, so that both sides were spared the acrimony, stress and expense of litigation. But secondly, the offer was made to protect the appellant’s position on costs.

Years, and after many hundreds of thousands of pounds later, all the respondents’ claims failed. The appellant had acted sensibly and proportionately at the outset; the respondents had not. In that context, I note that the only claim which might even arguably have been said to have been untainted by the respondents’ changes of position, and/or fundamentally flawed in principle (Footnote: ³), was the inspection claim, which at its highest was worth just £20,000 (namely less than the appellant’s Part 36 offer in any event).

Accordingly, in the particular circumstances of this case, I consider that the respondents’ failures to accept and then to beat the appellant’s Part 36 offer was a separate and stand-alone element of their conduct which was out of the norm, separately justifying an award of indemnity costs or, in the alternative, justifying such an order, when taken together with the nature of the claims pursued by the respondents.

On one view, it might be said that that conclusion should lead to an order for indemnity costs dating back to March or April 2015. But, taking into account all the circumstances of the case, it is I think appropriate to limit the indemnity costs to the period after they had had time to digest the Court of Appeal judgment of April 2017 (Footnote: ⁴). It was unreasonable beyond any doubt that the respondents did not accept the Part 36 offer once they knew that their omissions case was not open to them. Accordingly, the answer to the question formulated in paragraph 80 above is 7 May 2017.

I should stress that I have reached this conclusion by asking myself the questions that the judge should have been requested to ask himself, but was not, by reference to the particular facts of this case. I hope it is apparent from the previous paragraphs that I consider those facts to be relatively extreme. An order for indemnity costs was necessary and appropriate here because, on any view, this was a situation very similar to Excelsior: namely the pursuit of speculative/weak claims against the background of an offer that was unreasonably refused and subsequently not beaten.

The figure produced by an approved cost budget mechanism (CPR r.3.12-r.3.18) is a different thing to the final assessment of costs following the trial. The former is prospective; the latter is retrospective. True it is that, in many cases, the approved costs

If there is an order for indemnity costs, then prima facie any approved budget becomes irrelevant. In Denton and Others v TH White Limited [2014] EWCA Civ 906, Lord Dyson MR and Vos LJ said at paragraph 43:

A similar comment can be found in the more recent decision of Warby J in Optical Express Limited and Others v Associated Newspapers Limited [2017] EWHC 2707 (QB), a case where indemnity costs were ordered after a Part 36 offer had been accepted out of time. Warby J said at paragraph 52:

The absence of an overlap between the cost budgeting regime on the one hand, and an order for indemnity costs on the other, was explained in detail by HHJ Keyser QC (sitting as a judge of the High Court) in Kellie v Wheatley and Lloyd Architects Limited [2014] 5 Costs LR 854; [2014] EWHC 2886 (TCC). He said:

I respectfully agree with that analysis. In principle, the assessment of costs on an indemnity basis is not constrained by the approved cost budget, and to the extent that my obiter comments in Elvanite or Bank of Ireland v Watts suggested the contrary, they should be disregarded.

Before setting out briefly my reasons for rejecting Mr Oram’s submissions on this issue, I should say that, on detailed assessment, the figure for the appellant’s costs of not less than £724,000 odd is likely to be found to be unreasonable. I find it difficult to comprehend how such costs were incurred in a dispute about a garden in Highgate when the appellant’s original involvement lasted no more than a few weeks and was not the subject of charge. Accordingly, I am confident that, even on the indemnity basis from 7 May 2017 onwards, the costs finally determined on assessment are likely to be less than that figure.

The first answer to Mr Oram’s submissions is that I am not persuaded that there was a clear costs management order with an approved budget figure which is a reliable comparator with the amount of costs actually incurred. There is no r.3.15 order to that effect. The best that Mr Oram could do was to show particular figures for certain future phases which O’Farrell J approved on 1 December 2017 (paragraph 15 above). However, even those figures were uncertain because her order expressly envisaged that there would be revised costs budgets. On any view, therefore, there was no clear or settled starting point as envisaged by the costs management regime. Accordingly, I am not persuaded that this is a case in which it is appropriate to compare the figure of £415,000 with the much higher figure actually incurred.

Secondly, for the reasons explained in Section 8.2 above, there is as a matter of principle no overlap between a costs budget, which will have been approved on the basis of a projected series of figures for costs that were assessed as reasonable and proportionate, and the actual costs to be assessed by reference to the indemnity basis (where reasonableness might still be an issue, but proportionality is not). Thus, even if there had been an approved budget figure, it could not affect whether or not the court should make an order for indemnity costs.

In my view, because these were speculative/weak claims, and/or because the respondents unreasonably refused to accept the Part 36 offer that was made early, and which they then failed to beat, this is an appropriate case for indemnity costs. In my view, the judge was not assisted by counsel then appearing, and so did not ask himself the right questions on these issues. That was the reason why he fell into error. The discretion therefore falls to be re-exercised in the way set out above.

If my Lady and my Lord agree, I would order indemnity costs in favour of the appellant from 7 May 2017, being one month after the Court of Appeal judgment. The appellant’s costs prior to that date will be assessed on the standard basis.

I agree.

I also agree.