Arkin v Borchard Lines Ltd & Ors No.3

At the end of a trial lasting 49 days I gave judgment dismissing the claim for breaches of Articles 81 and 82 of the Rome Treaty and ordered that Borchard should recover 90 per cent of its costs against the Claimant and that the 2-4 defendants – Camomile Lines, Ltd, Furness Withy (Shipping) Ltd and Manchester Lines Ltd – should recover 80 per cent of their costs against the Claimant and pay 8 per cent of the Claimant’s costs.

By orders dated 27 July 2001 and 10 August 2001 Borchard obtained permission to join as Part 20 defendants Zim Israel Navigation Company (“Zim”), KNSM, DNOL as well as the existing defendants, Furness Withy and Camomile Lines. Borchard’s application for joinder was made about four months before the date then fixed for the trial. There were also other Part 20 defendants who took no part of the trial. The basis of this multiple joinder was that each of the Part 20 defendants was a co-participant in one or both of the conferences at the material time and that each was under an obligation as such or as a matter of law to contribute to or indemnify Borchard if it were held liable to the Claimant. However, Borchard maintained its claim against the Part 20 defendants.

However, the joinder of Zim was brought about for additional reasons. In particular, the Claimant had made specific allegations as to Zim’s conduct as amounting to an abuse of the dominant position of the conferences on the UK and North European markets. These allegations related specifically to Zim’s conduct on the South Africa – Israel route, the case put forward being that Zim’s conduct in that so-called ancillary market was an emanation of the conferences’ policy of pressuring Mr Arkin’s line into withdrawal from the UK and North European markets. Additionally, it was clear from an early stage in the pre-trial hearings that evidence of central importance to the case against all the other defendants was likely to be in the possession of Zim. This evidence consisted of the minutes of and working documents relating to the FMC committee and the oral testimony of the person who appeared to have played a central part in its operations, Mr Levy, an employee of Zim.

It was against this background that the court raised the question of how the trial could be sensibly conducted in the absence of Zim, which was outside the jurisdiction and whose employees could not be forced to give evidence. A strong indication was given that Zim ought to be joined. Leave was given to Borchard to serve Zim outside the jurisdiction as a necessary or proper party. The 2-4 defendants did not apply for such leave

Borchard’s application to join Zim has been criticised by Zim on the grounds that Borchard failed first to attempt to engage the co-operation of Zim in providing documents and witnesses. Whereas it is true that Borchard did not do so and, indeed did not even send a letter before application to serve out, I am satisfied that there was nothing intrinsically unreasonable in joining Zim. In particular, Zim occupied a central position in Mr Arkin’s allegations of abusive conduct and, on the face of it, had evidence directly material to those allegations. Secondly, although Zim had given a measure of co-operation to Ms Holmes of the 2-4 defendants’ solicitors in the course of her visit to Israel in 2000, it was far from clear whether they had approached the process of disclosure of documents as effectively and searchingly as would have been the case if they had been a party to the proceedings. Thirdly, it was quite unrealistic for Borchard to fight the claim on the basis that if Mr Arkin succeeded, separate proceedings could be pursued against Zim. The risk of inconsistent findings was far too great to leave that to chance.

In the event, Zim played a major part in the trial. This it was clearly entitled to do. It is absolutely axiomatic that a Part 20 defendant is entitled to protect its position by fighting the main action by its own evidence, provided that, in the interests of good case management, it does not unnecessarily take up trial time and cause increased costs by duplication of evidence already adduced by the other parties. There was no such duplication in this trial. Indeed, Zim adduced evidence of facts, disclosed documents and made available factual witnesses – Mr Levy and Mr Stramer who gave oral evidence and Mr Simkin and Mr Apel by written statements – who gave important evidence helpful to the court.

The 2-4 defendants did not join Zim as a Part 20 defendant. Their legal advisers may have taken the view that, even if Zim were not joined, it would have made its documents and employees available as ultimately it did following joinder.

Possibly they foresaw the potential exposure to costs which is the basis of the dispute now before me.

That dispute, between Borchard on the one hand and Zim and the Part 20 defendants represented by Davies Arnold Cooper (“DAC”) – KNSM and DNOL who became parties to the proceedings only because Borchard joined them, arises from the fact that Mr Arkin has no resources to satisfy the costs orders that have been made against him and, following the judgment in Arkin v. Borchard (No.2), no costs order has been made against MPC as a non-party funder of Mr Arkin’s litigation costs. It is therefore virtually inevitable that none of the defendants will be able to recover any part of the costs orders already made in their favour. Consequently, given that, with the dismissal of Mr Arkin’s claim against Borchard, the latter’s Part 20 claim against Zim also had to be dismissed, if Borchard were now ordered to pay Zim’s costs and left to recover those costs in addition to its own costs against Mr Arkin, the burden of Zim’s costs would ultimately fall on Borchard because recovery from Mr Arkin would be impossible. On the other hand, if I were not to order that Zim must look exclusively to Mr Arkin for its costs, the burden of those costs would fall on Zim just in the same way as each of the other defendants will have to bear its own costs.

Mr Peter Irvin has strongly argued on behalf of Borchard that it would be deeply unfair to Borchard if it were left with the burden of Zim’s costs. Zim was as much a conference participant as Borchard and other defendants and Mr Arkin could well have chosen to join Zim as co-defendants notwithstanding they were out of the jurisdiction. Borchard therefore reasonably took the course of joining Zim and thereby benefited the other defendants, who together with Borchard, gained access to Zim’s evidence. Zim argues that the usual order for Part 20 costs where the main action claim fails is that the Part 20 defendant recovers them from the successful defendant. It was Borchard’s choice to join Zim, which was unnecessary in all the circumstances and Borchard must have appreciated that it risked having to bear Zim’s costs if its defences against Mr. Arkin succeeded. Indeed, it was well-known to all parties at all times that Mr. Arkin was without significant resources and that he was having “a free ride” as regards the defendant’s costs. Further, Zim adduced expert evidence on shipping industry economics from Professor Yarrow, which was of assistance to all the other defendants, including Borchard. Indeed Zim argue that because Borchard chose to call no expert evidence, but instead to rely on the expert evidence adduced by the 2 – 4 defendants and Zim, Borchard should bear the burden of Zim’s costs arising out of Mr. Arkin’s impecuniosity.

On behalf of those Part 20 defendants whom he represented Mr. Steven Gee QC argued primarily in support of Zim’s position, contending that Borchard should be ordered to pay the costs of KNSM and DNOL thereby leaving it to Borchard to attempt to recover those costs from Mr Arkin. However, Mr. Gee also argued for an additional direct costs order in favour of his clients against Mr Arkin which was intended to recover such part of their costs directly as Mr Arkin could satisfy and the balance of their costs from Borchard. It was said that if Borchard became insolvent there was a risk that such costs as it was able to recover from Mr Arkin would never be transmitted to the DAC Part 20 defendants who would have to prove for them in competition with Borchard’s other creditors. Finally, as an alternative approach, Mr. Gee submitted that because the defence of all the defendants to Mr Arkin’s claim depended heavily on the expert evidence adduced by the 2-4 defendants and the Part 20 defendants, including Zim, all of whom were conference members directly or indirectly exposed to Mr Arkin’s claims, and as such were “joint venturers”, it would be fair and equitable for each of Borchard, Zim and the 2-4 defendants and Part 20 defendants represented by DAC to bear a share of the total costs of expert evidence, leaving other costs to be the subject of conventional costs orders.

Part 20 claims of the kind advanced by Borchard are treated by the CPR as they were by the RSC as in the nature of separate proceedings but which are, as a matter of proper case management, often appropriately ordered to be tried at the same time as the main action upon which they are based. Thus, for example, a Part 20 defendant is not entitled to security for costs against the claimant in the main action: see my judgment dated (2^nd August 2002) because the Part 20 defendant is not a party who is a defendant to the proceedings brought by the claimant for the purposes of CPR 25.12. Accordingly, if the Part 20 claimant’s claim in the Part 20 proceedings is dismissed following his successfully defending the main action claim, the Part 20 defendant is entitled to his costs against the Part 20 claimant in accordance with the underlying principle that as a general rule costs follow the event (CPR 44.3 (2) (a).). Since the Part 20 claimant will usually have acted reasonably in joining the Part 20 defendants and since the claimant in the main action must be taken to have foreseen the eventuality of the defendant’s joinder of the Part 20 defendant and his liability in costs to the Part 20 defendant should the claim in the main action fail, the normal order for costs includes an order that the defendant should be entitled to include in his costs as against the claimant those costs of the Part 20 proceedings for which he has become liable to the Part 20 defendant. Such costs orders are entirely consistent with the conceptual basis of the general rule under CPR 44.3(2)(a), namely the public policy objective of deterring the bringing of ill-founded defences: see Roache v. News Group Newspapers Ltd. (1992) [1998] EMLR 161 at page 166, per Sir Thomas Bingham MR. The fundamental purpose of this rule is thus deterrent and not compensatory, although its implementation is effected by indemnification of the successful party to the extent of his reasonably incurred costs.

Where the claimant, having advanced related claims against two defendants on alternative but related grounds, succeeds against one but loses against the other, the courts have long adopted an approach to costs orders permitting the successful defendant to recover his own costs either against the claimant or against the losing defendant, in the former case permitting the claimant to pass on the successful defendant’s costs to the unsuccessful defendant.

The development of these alternative orders was explained by Sterling LJ in Sanderson v. Blythe Theatre Company [1903] 2 KB 533. In summary, the jurisdiction to make costs orders in favour of the successful defendant against a claimant whose claim fails against him but succeeds against another defendant was seen to give effect to the rules of court which required that in jury trials costs must follow the event unless the trial judge for good reason otherwise ordered. Where the trial was by judge alone, his discretion as to costs was wide enough to enable him to order that the unsuccessful defendant should pay the costs of the successful defendant. This latter form of order had been adopted in the Court of Chancery prior to the Judicature Acts when that court had power to permit the joinder of co-defendants and the Courts of Common Law did not and where the action was brought by the Attorney-General. Because a costs order could not be made against the Attorney-General even if his claim were unsuccessful, it was considered appropriate to order the unsuccessful defendant to pay the costs of the successful defendant direct, thereby avoiding the injustice of the unsuccessful defendant not having to pay costs incurred by the successful defendant: see Attorney-General v. Corporation of Chester (1851) 14 Beav. 338. Following the Judicature Acts, the Court of Appeal in Rudow v. Great Britain Mutual Life Assurance Society (1879) 17 Ch. Div. 600 extended this cut-through form of order as appropriate to all cases of co-defendants. In Sanderson, supra, at pages 542-543 Stirling L.J. referred to the judgment of Sir George Jessel MR in Rudow and to his comments on the previous procedure of the Court of Chancery in ordering the unsuccessful defendant to pay to the plaintiff the successful defendant’s costs:

“Jessel M.R. said that the machinery which had been adopted by the Court of Chancery was no longer either necessary or useful, but was mischievious. I understand this to mean that it was not necessary, because under the Judicature Acts and rules the Court had full power to order one defendant to pay costs to another, and that it was mischievous because the plaintiff who was ordered to pay costs in the first instance, and then to have them over against a defendant, ran the risk of losing them if the defendant proved to be insolvent. The change of practice ought to be adhered to wherever it is practicable to do so. It appears to me that this practice is not applicable to a case in which the judge holds that there is not good cause to deprive a defendant of costs against the plaintiff under the last clause of Order LXV., r.1, and yet is of opinion that the costs so occasioned ought properly to be borne by another defendant. In that state of things it seems to me that, for the attainment of justice, recourse may and ought to be had to the old practice.

It is to be observed that when the Court of Chancery ordered the costs of one defendant to be paid by the plaintiff and added to his own, and the whole to be borne by another defendant, the Court was not giving substantive relief by way of indemnity or damages or otherwise, but was simply exercising its jurisdiction over the costs of a pending suit. This particular form of order was, as Jessel M.R. calls it, “machinery” for properly disposing of the costs, and this is shown by the change in the form of order when, as in the case of the Attorney-General, the plaintiff could not be ordered to pay costs. Having regard to this and to the express directions of Order LXV, rule 27, regulation 37, I think that the term “costs” in s.5 of the Judicature Act, 1890, ought to be read as including costs which a particular party may be ordered to pay in the first instance and then to have over against another.”

Thus, for over a century, in cases where a claimant succeeds on alternative claims against one of two defendants and fails against the other, it has been regarded as an appropriate procedure to make a cut-through order and the justification for such orders has been recognised as avoiding the unfairness to the successful defendant that might arise if an order that routed his costs from the unsuccessful defendant through the claimant could not be effectuated because of some characteristic of the claimant, including, for example, his insolvency. Thus, although alternative claims against different co-defendants are brought in the same proceedings and do not involve separate third party proceedings against those parties, it is worth noting for present purposes that the incidence of insolvency has been treated for at least a century as relevant to the way in which the court exercises its discretion as to the machinery to be adopted to give effect to its orders in respect of the incidence of costs.

In Bankamerica Finance Ltd v. Nock [1988] AC 1002 where the hirer of a car from the claimant finance company discovered that it was stolen and terminated the hiring agreement and the finance company then sued the hirer for damages for wrongful repudiation of the hire agreement and in the alternative the dealer from which it had purchased the car for breach of the sale contract and where the hirer succeeded against the claimant and the claimant succeeded against the dealer who was insolvent, the trial judge had ordered that the dealer should pay part of the costs of the hirer direct as well as the costs of the claimant finance company, declining to make an order routing all the hirer’s costs through the finance company. The consequence of his order was that neither the hirer nor the finance company were able to enforce their costs orders against the dealer except by proving in the dealer’s liquidation, but the hirer was able to recover from the finance company that part of his costs for which he had an order against it. In addition, the finance company would have to prove in the dealer’s liquidation for its damages. It had been contended on behalf of the hirer that the judge should exercise his discretion in favour of an order that the hirer recover his costs from the claimant finance company, which would then be left to recover what, if anything, it could from the insolvent dealer.

The House of Lords (Lord Brandon) held that it was a proper exercise of the judge’s discretion to make the costs orders which he had made. At pages 1011 to 1012 Lord Brandon said this:

“With regard to the third question, if the dealers had not been insolvent, it would, as I pointed out earlier, have made no difference in the end which of the two forms of order was made. The real question is therefore whether, having regard to the fact that the dealers were insolvent, the judge could not, if he exercised his discretion judicially, have made a Sanderson order, but was bound to make a Bullock order. As to this it is true that, if the Sanderson order was made, the hirer would probably have to bear his own costs. He would, however, recover the sum of £8,344.83 awarded to him against the finance company, whereas the finance company would probably be unable to recover from the dealer either the sum of £23,996.34 awarded to it or its own costs. There would then be hardship to both parties, but more to the finance company than to the hirer. By contrast, if a Bullock order was made, the hirer would recover from the finance company both the sum of £8,344.93 awarded to him and his costs. The finance company on the other hand would probably recover nothing: neither the sum of £23,996.34 awarded to it, nor its own costs, nor the hirer’s costs which it would have to have paid. All the hardship would then be to the finance company and none at all to the hirer.

The judge must have been aware of these matters. Having regard to them it seems to me impossible to say that the judge could not, in the judicial exercise of his discretion, have made a Sanderson order but was bound to make a Bullock order. On the contrary the balance of hardship seems to me, not to require the judge to make a Sanderson order than a Bullock order, but at least to provide a legitimate ground for him, in the judicial exercise of his discretion, to do so.

In this connection it is pertinent to observe that in Rudow v. Great Britain Mutual Life Assurance Society (1881) 17 Ch D 600, 607-608 Jessel MR expressed the view that, in the case of the present kind, the established practice was always to make a Sanderson order rather than a Bullock order. I recognise at once that this extreme approach must be regarded today as going too far. But the fact that it was adopted formerly is a strong indication that, while a Sanderson order cannot be mandatory, a Bullock order cannot be mandatory either.”

Here again, therefore, in a case involving co-defendants, the insolvency of one of the parties was treated as a consideration which could justifiably be taken into account in exercising the court’s discretion as to the appropriate order. Further, as Lord Brandon observed, in such cases neither a Sanderson order nor a Bullock order could be regarded as mandatory. The form of order was in the discretion of the court. The exercise of that discretion could be based on a consideration of the “balance of hardship” which would be suffered as between the two forms of order.

That being the approach of the courts in cases involving successful and unsuccessful co-defendants, I must now consider the cases involving third party and Part 20 proceedings.

In Johnson v. Ribbins and others [1977] 1 WLR 1458 the facts were that the plaintiff co-mortgagor commenced proceedings against the defendant mortgagee alleging a sale of the mortgaged property at an undervalue. The defendant joined the estate agents who had advised on and handled the sale as third parties. After this joinder the plaintiff applied for and obtained legal aid. That brought into operation section 7(6) of the Legal Aid Act 1974 by which it was provided that:

“the rights conferred by this Part of this Act on a person receiving legal aid shall not affect the rights or liabilities of other parties to the proceedings or the principles on which the discretion of any court or tribunal is normally exercised.”

The trial judge, having dismissed the plaintiff’s claim and therefore also the claim by the defendant against the third party, taking into account the fact that the claimant was legally aided stated that he was adopting the principle that the impact of legal aid should fall on the defendant and third party alike and he therefore ordered that the plaintiff should pay the costs of the defendant and of the third party direct, thereby avoiding the position where the defendant was left to pay costs to the third party which he could probably not recover from the plaintiff. The judge also adopted an alternative basis for such exercise of his discretion by stating that if there were no such principle relating to a legally-aided party then it would be right in that case “to produce equality of sacrifice”.

The court held that section 7(6) precluded the judge from taking legal aid into consideration in deciding how he should exercise his discretion. Even if, on his alternative basis, the decision was not in conflict with section 7(6), the objective of producing equality was still misconceived. Goff LJ., delivering the judgment of the court, observed at pages 1463-1464:

“In the circumstances of this case it is abundantly clear that the real and only fight was between the plaintiff as the alleged owner by adverse possession and the true owners, the third parties, and, accordingly, we should have been prepared to order that the plaintiff should pay their costs directly. However, the defendants’ notice of appeal only asks that they may be at liberty to add the costs which they have been ordered to pay to the third parties to the costs which the plaintiff should pay to them.”

In the exercise of that discretion, however, in our judgment, the court

“should be guided by the principle that normally costs follow the event as is expressly provided by RSC Ord 62, r.3(2) and should, therefore, normally order the defendant, though successful in the action, to pay the costs of the third party if he also be successful. Then if in the circumstances of the case these costs ought fairly to be borne by the plaintiff the court will further order that they be added to the defendants’ costs of the action as against the plaintiffs.

Mr Balcombe argued that one cannot apply Ord 62, r3(2) to third party proceedings, because the only relevant event is the outcome of the proceedings as a whole. One cannot, so he submitted, look at the third party proceedings separately. We cannot accept that view as a general proposition and certainly not as applicable to this case, though it might be true in particular cases. Generally, we think it will be found, and certainly so in this case, that the defendant and the third party stand in relation to one another as if the defendant had brought a separate action against the third party, and in our judgment, this conclusion is fortified by the Judicature Act 1925, section 39(1)(b) and the note 16/1/17 in The Supreme Court Practice (1976) vol 1, p230.”

Later in his judgment he said this at page 1464:

“Thus, the question resolves itself in our view into this, namely, is there on the facts of this case anything which should lead the court in exercising its discretion to depart from the normal principle that costs follow the event? We can see nothing. On the contrary, in our judgment, the facts call strongly for it to be observed.

Apart from the impact of legal aid the consideration of which, as we have already observed, is excluded by the Act itself, we can see nothing which the defendants can call in aid except the impecuniosity of the plaintiff, but it cannot be right to deprive a third party of an order for costs to which he is otherwise entitled against the defendant, because the defendant when looking to the plaintiff for reimbursement finds a person not worth powder and shot.”

Finally, at pages 1465-1466 Goff LJ observed:

“For these reasons, in our judgment, the order made by the judge was not the correct one, but it is said that whilst he was, as he expressly stated, consciously and deliberately adopting a principle which we have held to be wrong we cannot interfere because he went on to reach the same conclusion as a matter of judgment. Thus, it is said he exercised his discretion and in doing that made no error in principle, and so we are bound by what he did and cannot substitute our discretion for his.

In our judgment, however, that is not so. Even assuming that he was not, in exercising his discretion, wrongly taking into account the impact of legal aid, still, in our judgment, he erred in principle because he paid no attention to the normal rule that costs follow the event, and if he was not being influenced by considerations of legal aid then he was incorrectly depriving the third parties of their prima facie right to an order for costs against the defendants on the ground of impecuniosity of the plaintiff. That was an error in principle.”

Further points relied upon by the judgment to justify the maintenance of separate proceedings were the long-winded manner in which the claimant’s case was presented and the failure of the defendant before joining the third party to invite it to admit liability conditionally on the findings in the main action.

I would observe that this judgment did not refer to the width of the discretion on costs identified by such cases as Rudow and Sanderson where co-defendants were involved. It is not clear whether the court was referred to what had been said in either of those cases. Further, the facts are not unusual and the court’s view was clearly that unless there were very unusual circumstances a cut-through costs order should not be made, regardless of the incidence of legal aid, as a device for ameliorating the impact of the impecuniosity of the claimant. That could only be because it was the court’s view that where there were separate third party proceedings costs in those proceedings should normally follow the event as provided in RSC order 62, rule 3(2) and that the impecuniosity of the claimant was not in itself a feature which was sufficient to justify departure from the normal rule.

In Taylor v. UKF Fertilisers Ltd (CA) (Unrep – 1 November 1988) a claim for personal injuries had been made by an employee against the defendant employers who in turn had joined as third parties the company from whom they had hired the fork lift truck off which the claimant had fallen. It was alleged that the truck was unsafe. The claimant’s claim failed. He was legally aided. The judge ordered him to pay the costs of the defendant employers alone totalling £450. He declined to make any order for costs in the third party proceedings. His explanation for declining to do so was that at the trial the third party had taken an identical position to that of the defendant, including calling expert evidence to refute that adduced by the claimant. The judge said that therefore a just decision on the third party costs was that each of the defendant and third party should bear their own costs. In the Court of Appeal Taylor LJ, with whom O’Connor and Nicholls LLJ. agreed, having expressed the view that this was not a sufficient reason for departing from the normal rule that costs should follow the event, as expressed in Scherer v. Counting Instruments [1986] 2 All ER 529, addressed an argument based on Bankamerica Finance v. Nock, supra, in the following passage:

“Mr Gilmour submitted, in an attempt to supplement the reasoning, that the learned judge must have had in mind the plaintiff’s impecuniosity and must therefore have tailored his order for costs so as to spread its effects evenly between the defendants and the third party. That approach was upheld by the House of Lords in Bankamerica Finance Ltd v. Nock, but I do not think that one can imply that reasoning here. Firstly, the learned judge, when specifically responding to a request for his reasons, made no mention of it. Secondly, by ordering the plaintiff to pay costs to the defendants, albeit only £450, and leaving the third party to pay their own, he was not treating them equally. Accordingly, I am driven to the conclusion that the learned judge in exercising his discretion did not base it on any sufficient reason and was therefore not exercising it judicially. It follows that, despite his refusal of leave, this court has power to entertain an appeal against his order.

It is possible to read this passage as a somewhat oblique indication (certainly obiter) that achieving equality of treatment might have amounted to a sufficient reason for the judge’s order had he expressed that as his purpose and had his order achieved equality.

It is however clear from the following passage in the judgment that Johnson v. Ribbins was cited to the court.

O’Connor LJ said this:

“It will be seen in the Bank America case that a successful party was effectively deprived of costs because, on the justice of the case, the unsuccessful party, who was also deprived of both costs and remedy from a co-defendant who was insolvent, would suffer a greater injustice, and although the making of a Sanderson order in that case was considered to be unjust and unappealable in the Court of Appeal, Lord Brandon in the House of Lords took a different view as to the justice of the case and supported the judge’s finding.

It might be that in certain circumstances, where a third party is joined in an action such as this, that the judge could make an order which would result in a successful third party not recovering costs because a successful defendant was also not recovering costs. I have considered whether it is possible to construe the judge’s reasons as falling within the Bank America principle, but I find myself, like Lord Justice Taylor, quite unable to do so. It would have been a simple thing to state and it would have resulted in an entirely different order from that which the learned judge made. So I am satisfied that this, in effect, is a case where there has been no exercise of discretion by the judge, because that is the effect as he purports to have exercised it for reasons which cannot be supported.”

If the decision in Johnson v. Ribbins, supra, is read as imposing an inflexible principle that where there are third party proceedings it is unjustifiable to depart from the general rule that costs follow the event in those proceedings on the grounds that due to the impecuniosity of the claimant the defendant will be left with the burden of the third party’s costs, then it is certainly not consistent with the well-established approach to the incidence of costs as between co-defendants where one successfully defends the claim but the claimant is insolvent. When the court makes a cut-through order enabling the successful defendant to recover its costs direct from the unsuccessful defendant, it departs from the general rule that costs must follow the event for in such a case there is no order for the claimant to pay the successful defendant’s costs. The position therefore would be that, whereas the discretion to depart from the general rule can be exercised with reference to the impecuniosity of the claimant where co-defendants are involved, there could be no such justification in relation to third party proceedings. But such a distinction could be maintained only if there were some characteristic of third party proceedings which as a matter of policy normally distinguished them from proceedings involving co-defendants and therefore called for the exclusion of that consideration.

There is no doubt, in my judgment, that third party proceedings, as much under the CPR, as under the former RSC, are treated as separate proceedings from the main action. It follows that if the impecuniosity of the claimant in the main action is to carry weight in the discretionary disposition as to costs, that would be to introduce into the third party proceedings consideration of an extraneous circumstance affecting a party who played no part in the third party proceedings. Although it is true that both sets of proceedings may be tried together, that is a matter of good case management and does not detract from the essentially separate nature of the two proceedings. The fact that it is the policy of the law to give effect to access to justice permitting an impecunious claimant to sue and so to expose an ultimately successful defendant to shoulder the burden of his own costs see Hamilton v. Al Fayed (No.2) [2002] 3 All ER 641, per Simon Brown LJ at 658-659, raises the question whether, if that defendant, with knowledge of the claimant’s lack of resources, then chooses to protect his position by joining a third party Part 20 defendant, it is in the interests of justice that the third party should be placed in a similar position as to the recovery of his costs, if successful, to that of the defendant who has chosen to join him. It is indeed hard to see why the apparent injustice to the defendant by reason of his inability to recover his costs from the impecunious claimant should ordinarily be spread to the third party who has been sued involuntarily by a sufficiently resourced defendant. If, as it is ultimately decided, he has been wrongly sued and is under no liability to the defendant, to expose him by a costs order to the impecuniosity of the claimant would normally not only be unnecessary to do justice to the defendant but would do serious injustice to the third party. This, as I see it, represents a quite different position from that which arises where there are co-defendants both of whom have been sued by the same impecunious claimant. In such a case there is often a far stronger character of injustice if the successful defendant is deprived of his costs which can in appropriate cases be cured by a cut-through or Sanderson order causing them to be paid by the unsuccessful defendant. The reference by Goff LJ. to the decision of Diplock J. in L E Cattan Ltd v. A Michaelides & Co [1958] 1 WLR 717 at 720 in relation to the arbitration of chain contract disputes suggests that the maintenance of the essential separability of third party proceedings carried fundamental weight in the decision in Johnson v. Ribbins, supra.

I therefore approach the exercise of my discretion on the basis that it will only be in exceptional cases that the separability principle will justifiably be departed from. Although in Taylor v. UKF Fertilizers, supra, O’Connor L.J. clearly envisaged that in “certain circumstances” a cut-through order effectively depriving a successful third party of his costs might be justifiable, it is also clear from the judgment of Taylor L.J. that neither the insolvency of the claimant nor the mere common interest of the third party and defendant in the outcome of the main action and their virtually identical position in challenging the claimant’s case are when taken alone amongst the special circumstances that render the case exceptional.

Further, in the present case, the Part 20 proceedings were not such as would necessarily be conclusively determined by the result of the main action. I can see that in cases where, if the defendant lost to the claimant, it would inevitably follow that the third party must be liable to the defendant, to impose on the defendant the burden of the Part 20 defendant’s costs as well as his own might amount to an injustice so great as to justify making an order that the Part 20 defendant should recover his costs direct from the impecunious claimant. However, where, as in this case, there were likely to be discrete Part 20 issues arising out of the Conference Agreements and the conduct of Zim in relation to conference members, the Part 20 issues do no more than overlap on the issues in the main action. They are not co-extensive.

Additionally, this is not a case where Borchard and the third party Part 20 defendants made common cause as to joinder. Quite the contrary. Borchard did not send letters before action or invite conditional acceptance of liability before commencing the Part 20 proceedings. Instead it pursued an arms length approach to the Part 20 defendants which was consistent with the maintenance of the separate nature of the Part 20 proceedings. Further, Borchard has derived from the joinder of Zim the benefit of both factual and expert evidence, while adducing no expert evidence itself. It may well be that even if Zim had not been joined, the factual evidence would still have been available. However, the expert evidence would not.

In these circumstances this would not, in my judgment, be an appropriate case in which to make a cut-through order confining the Part 20 defendants to recovery of their costs direct from Mr Arkin.

Should there be an order whereby the costs borne by the Part 20 defendants and the main action defendants should be aggregated and divided rateably between those parties and Borchard on the assumption that Mr Arkin will never be able to pay them? The justification for such an order would be that all the issues of expert evidence had to be decided for the common benefit of all the defendants and Part 20 defendants so that it would be unjust if Borchard, due to the impecuniosity of Mr Arkin, had the whole burden of those costs incurred by Zim and the other Part 20 defendants and the DAC defendants.

As to this alternative submission, I am unable to accept that to make an order imposed on Borchard any share of the costs of expert evidence to the DAC defendants would be appropriate. Cost-sharing between successful co-defendants may be something achievable as a matter of agreement, but it is not normally appropriate that it should be imposed by the court. The ordinary course is for co-defendants to co-operate in the provision of expert evidence and to agree to share the costs. In the present case there was no such agreement and there is no basis upon which the court can retrospectively impose such a sharing arrangement. The costs must lie where they fall in so far as they are irrecoverable from Mr Arkin.

In relation to the more limited sharing of the aggregate of expert evidence costs between Borchard, Zim and, to the extent to which they have incurred them, KNSM and DNOL, the question is whether to require Borchard to pay those costs in full, leaving it to its almost certainly fruitless recourse to Mr Arkin, would give rise to such substantial injustice as to override the general rule that the costs must follow the event and that the separability of the Part 20 proceedings should be maintained. Such an order would involve KNSM and DNOL and Zim being entitled to recover from Borchard only a proportionate share of those expert costs and bearing the balance themselves.

In all the circumstances I have come to the conclusion that my discretion should not be exercised in this way. In reaching this conclusion I have very much in mind that Borchard was aware at the time when it joined the Part 20 defendants that it was facing an impecunious claimant and that absent a section 51(3) order against an outside funder (MPC) it was exposed to the risk of the court taking the approach to costs identified in Johnson v. Ribbins. It made no attempt before the trial to co-operate with the other conference members as to the provision of or the cost of expert evidence and was content to pursue its defence by the cheapest means possible – reliance on the expert evidence adduced and paid for by others. There was a very low level of co-operation between Borchard and all the other parties as to how expert evidence was to be deployed by way of defence.

Taking all these matters into account I am not persuaded that this is an appropriate case in which to make such an unusual costs order. Nor do I accept that there is any justification for a hybrid order of the kind referred to by Mr Gee on behalf of KNSM and DNOL. There is no cogent evidence that Borchard does not have the resources to satisfy the costs bill and, even, if there were, such an order would not be justified in view of the approach of the courts to the incidence of insolvency of a claimant referred to at paragraph 33 above.

In the event, the Part 20 defendants are entitled to orders for costs against Borchard. Those orders should reflect the orders made against Borchard in the main action to the effect that Zim will be entitled to 90 per cent of its costs from Borchard and the DAC Part 20 defendants 80 per cent of their costs. The DAC Part 20 defendants were represented by the same solicitors and counsel as represented the 2-4 Defendants and, in the absence of their express dissociation from all those defendants’ submissions, it is just that they should recover costs from Borchard reduced to the same extent as the 2-4 Defendants’ costs were reduced as against Mr. Arkin. Borchard will be entitled to add those costs to its costs recoverable from Mr. Arkin. The order for Mr. Arkin to pay 90% of Borchard’s costs will of course impact only on the costs incurred by Borchard itself and not on the component of already-reduced costs payable by Borchard to the Part 20 defendants.