BGHZ 55, 176 VIII. Civil Senate (VIII ZR 61/69) Jungbullen-decision
11 January 1971
Translated by:
Mr Raymond Youngs
Professor B.S. Markesinis

The defendant is an attorney who represented the plaintiff’s father in divorce proceedings. In January 1972 the plaintiff’s father and mother met in the defendant’s office, where they signed a divorce agreement drawn up by the defendant. It contained the following clause:

§ 6. As to the house, the parties agree that the half belonging to Mrs M is to be transferred to the three children in equal parts. Mr M hereby agrees not to sell his half but to transfer it to his present legitimate children. An appropriate notarial contract to this effect is to be concluded immediately after the divorce is final. Mr M further promises that once the divorce is final he will indemnify Mrs M against any liabilities arising from the house or its construction. . . .

A divorce decree was granted in February 1972, and the defendant, in the name of the plaintiff’s father, thereupon waived any rights of appeal, as did the mother’s attorney. The plaintiff’s mother now refuses to transfer her interest in the property to the plaintiff and his siblings.

The plaintiff claims damages for breach of the defendant’s duty as attorney. The Landgericht rejected the claim but the Oberlandesgericht allowed it. The defendant was permitted to appeal, but his appeal was dismissed.

I. [the reasoning of the Court of Appeal.]

II. Despite the appellant’s contentions, this reasoning is sound in law.

1. [The defendant was in breach of his duty as attorney.]

2. Nor is there anything wrong in law with the Court of Appeal’s holding that although there was no contract between the plaintiff and the defendant, the plaintiff could sue the defendant for damages for his faulty breach of contract.

(a) The Court of Appeal found that there was here a contract with protective effect for third parties and that the plaintiff’s claim arose therefrom. We do not have to decide whether this is so.

(aa) Certainly an important factor pointing in that direction is that the plaintiff was the son of the attorney’s client and was entitled to care and protection from him (compare BGHZ 61, 227, 233). The usual problem in cases of contracts with protective effects for third parties is whether the victim was someone the debtor could expect to be harmed by a breach of the contract. That is not the problem here. The very words of § 6 of the divorce agreement drawn up by the defendant show that the children were its sole beneficiaries, the only people apt to suffer if the agreement proved invalid.

The only question here is how far the protective effect of this contract works in favour of the children, in particular whether they have any claim for damages for breach of contract in their own right. Now the contract between client and attorney is such, given its nature and structure, that it can only be very seldom, whether one interprets the contract extensively or invokes § 242 BGB (see BGHZ 56, 269, 273; NJW 1975, 977), that the duties it generates can be sued on by third parties, for the fiduciary relationship between client and attorney makes it strongly bilateral and self-contained [references omitted]. Thus the fact that third parties have an interest in what an attorney does will not normally lead to any extension of his liability, even if those persons are named or known to him. However, an exception must be made where a contract drafted by the attorney is designed to vest rights in third parties specified therein, especially third parties who, as in the present case, are represented by the client. It is true that most of the cases where the courts have granted third parties a claim for damages arising out of a contract to which they were not parties have involved personal injury or property damage and its consequences (BGHZ 49, 350, 355; NJW 1955, 257; [other references omitted]), but it is not impossible for a third party to have a personal claim for economic loss caused by breach of subsidiary contractual duties (NJW 1968, 1929; BGH NJW 1975, 344). In drawing the line here one must certainly apply an especially stringent test: the circle of persons to whom the protective effect of a contract extends is to be narrowly drawn, so as to avoid blurring the line between contractual and tortious liability in an unacceptable manner (BGHZ 66, 51, 57; NJW 1974, 1189). It must always be borne in mind, in claims for purely economic loss, that the debtor is not to be made liable for the mere ricochet effect of his conduct on third parties.

(bb) Despite this, we cannot, on the special facts of the present case, fault the Court of Appeal’s holding that the plaintiff was drawn into the protective ambit of the attorney’s contract. The respondent invokes a decision of this court of 6 July 1965 (NJW 1965, 1955), but this is not quite in point. The court there did allow the daughter of a client to sue the attorney although she was not herself a party to the contract, but the court was reluctant to categorize the contract as one with protective effect for third parties [references omitted]. Contracts with protective effect for third parties are concerned with breach of subsidiary duties by the contractor (see BGH NJW 1975, 344), whereas in that case the question was really whether the attorney could be made liable towards the client’s daughter, the third party, for a breach of specific duties of performance [reference omitted]. Our case is clearly distinguishable.

(b) The plaintiff might also base his claim here on the concept of Drittschadensliquidation, a doctrine which borders on, if it does not actually overlap, the area of application of the doctrine of contracts with protective effect for third parties (see BGHZ 49, 350, 355). It would have been quite proper for the defendant’s client to indemnify his son, the plaintiff, for the harm he had suffered, and one could then infer from the fact that he brought suit as his son’s statutory representative that he was making an assignment of his own claim which the plaintiff, on the threshold of majority, could implicitly accept. But we need not pursue the matter here.

(c) In whatever legal or doctrinal category one puts the present litigated facts, the result must be that the plaintiff has a direct claim against the defendant attorney for compensation for the harm which he suffered as a result of the defendant’s failure to tell his father of the need to implement the agreement in § 6 of the divorce document. Any other conclusion would be inconsistent with the meaning and purpose of the attorney’s contract here and of the father–son relationship between the client and the plaintiff of which the defendant was well aware.

Addendum: White v. Jones

(Extracts from the transcript of the judgment of the Court of Appeal. A summary of the decision appeared in The Times of 3 March, 1993; the case has now been reported in [1993] 3 WLR 730).


The testator had two daughters: Carol White and Pauline Heath. When his wife died in January 1986 a family quarrel ensued as a result of which the testator proceeded to cut both his daughters out of his estate. In his will, executed in March 1986, he distributed his rather modest estate in various ways among three of his grandchildren and Mrs White’s first husband. Soon afterwards a reconciliation took place and the testator, regretting his rash action, informed his two daughters of the will and of his intention to change it. Both he and Mr Heath (a different son-in-law) contacted Mr Jones of the firm of defendant solicitors and gave both oral and written instructions for a new will under which the two daughters would inherit approximately two-thirds of the estate, the remainder being equally divided among all (five) grandchildren. Despite repeated communications, Mr Jones remained inactive though, in mid-August, and before leaving for his annual holiday, he did write an internal memo to a member of the firm’s probate division informing him of the need to prepare a new will. A proposed meeting with the testator on 17 September never took place since the latter died a few days earlier. The two daughters, having failed to reach an amicable settlement between them as to the way the estate should be divided, sued the solicitors. At First Instance, Turner J dismissed their action, holding that though the solicitors had committed a serious wrong, they owed no duty of care to the plaintiffs. A unanimous Court of Appeal (consisting of the Vice Chancellor, and Lord Justices Farquarson and Steyn) reversed the decision.

The vice chancellor . . .

The liability question

I start with some preliminary observations. First, the present case concerns the consequences in law of a solicitor’s failure to prepare a will. In my view there is no distinction in principle between this case and a case, such as Ross v. Caunters, where the solicitor’s failure lies in not warning the client about formal witnessing requirements. On this I cannot agree with the Judge. There is a difference of degree in that with the latter case the proposal to benefit the beneficiary has proceeded further. In the latter case the will has been drawn up by the solicitor and signed by the testator with the beneficiary named therein. In the former case the matter has not proceeded beyond the stage of the solicitor accepting instructions. It is possible the client might have changed his mind before executing the will. I do not think this can be a crucial difference when considering whether a liability in negligence exists, although it may be very material when considering whether a beneficiary has proved he has suffered loss through the solicitor’s negligence. In both cases the solicitor is in breach of his professional duty in carrying out his client’s instructions for the preparation and execution of a will, in the one case by doing nothing, in the other case by doing his work badly or incompletely. Hereafter I shall draw no distinction between these two cases, and for convenience I shall refer to the question whether in such cases a solicitor can be liable to an intended beneficiary as ‘the liability question’.

Knowledge of the testator’s intentions

Second, in the present case the prospective beneficiaries played some part in the steps taken by the testator towards the preparation of a new will. They were aware of his wishes, they were aware of his instructions to the solicitors, and they even acted as a channel of communication between him and Mr Jones. Those features are strictly irrelevant when considering the liability question. Let me explain why. There may be cases when a prospective beneficiary can show reliance on a solicitor’s advice in circumstances where a liability in negligence will clearly follow, on classic Hedley Byrne ([1964] AC 465) principles. For example, if a prospective beneficiary is present when the will is being signed and in response to a question the solicitor assures him it is in order for the prospective beneficiary’s wife to be a witness.

That is not this case. Here there is no question of any reliance of that nature. So I put ‘reliance’ cases on one side. Where reliance is present the law is not in doubt. Short of reliance, in my view awareness by an intended beneficiary of what the testator is doing cannot itself make the solicitor liable to the beneficiary if otherwise he would not be liable. Similarly in respect of assistance given by a prospective beneficiary to the testator in instructing the solicitor. It cannot be that if there are two intended beneficiaries, one of whom knows of the testator’s plans and his instructions to the solicitor and the other of whom does not, the negligent solicitor is liable to the former beneficiary but not the latter. That cannot be the touchstone of liability. The two cases must stand or fall together. Absent reliance as already mentioned, the solicitor must be liable in both cases or neither. In other words, the answer to the liability question must be applicable alike to the case where a prospective beneficiary is not even aware of his would-be benefactor’s intentions as to a case where he is.

Robertson v. Fleming

Third, Mr Matheson submitted that we are bound, by the House of Lords decision in Robertson v. Fleming (1861) 4 Macq. 167, to answer the liability question in favour of the solicitors. I do not agree. In that case all their lordships stated, or implicitly accepted, that in the absence of a contract between them a solicitor was not liable to an intended beneficiary. In rejecting the contrary view Lord Campbell LC (at p. 177) expressed himself forthrightly:

If this were law a disappointed legatee might sue the solicitor employed by a testator to make a will in favour of a stranger, whom the solicitor never saw or before heard of, if the will were void for not being properly signed and attested. I am clearly of opinion that this is not the law of Scotland, nor of England, and it can hardly be the law of any country where jurisprudence has been cultivated as a science. . . .

Lord Cranworth (at pp. 184–185) made an observation to the same effect.

I need hardly note that the law has moved on since 1861. This case was decided before Donoghue v. Stevenson [1932] AC 562, before Hedley Byrne, and before Murphy v. Brentwood. These twentieth century decisions of the House of Lords establish that there can be a liability in negligence in circumstances where the contrary view was regarded as axiomatic in the nineteenth century. In this regard Robertson v. Fleming does no more than apply the then established law on negligence to the particular case of a claim by a third party against a solicitor. What it does not do and cannot do is throw any light upon the answer to be given to the liability question by an application of the principles which, in the case of Donoghue v. Stevenson, have subsequently become part of the bedrock of the English law of negligence.

An open question

Fourth, neither in Caparo Industries Plc v. Dickman [1990] 2 AC 605 nor in Murphy v. Brentwood did the House of Lords expressly or implicitly doubt the correctness of Ross v. Caunters. In the Caparo case (at p. 636) Lord Oliver of Aylmerton noted that it gave rise to certain difficulties of analysis. However, in the Murphy case (at p. 486) he expressly left open the possibility that outside the reliance cases there may be cases, such as Ross v. Caunters, where there is a duty to take reasonable care to avoid pecuniary loss. In my view the position is simply that the House of Lords has left open the correctness of Ross v. Caunters; their lordships have expressed neither approval nor disapproval.

The effect of the solicitors’ argument

Fifth, in considering the liability question it is of the utmost importance to keep in mind that if there is no liability the result, as pointed out by Sir Robert Megarry V-C in Ross v. Caunters (at p. 303), is striking: the only person who has a valid claim against the solicitor has suffered no loss, and the only person who has suffered a loss has no valid claim. The executors can sue the solicitor for damages for professional negligence, but they can recover only nominal damages because the estate has suffered no loss. All that has happened is that, by reason of the negligence, on the testator’s death the estate passes to a different beneficiary. The intended beneficiary suffers a loss, but he has no right of recourse against the solicitor for the professional negligence which caused his loss. It would be a sorry reflection on English law if, indeed, that is the position today.


The House of Lords’ decision in Caparo v. Dickman established that, for there to be a duty to take reasonable care to avoid causing damage of a particular type to a particular person or class of persons, three factors must coalesce: foreseeability of damage, a close and direct relationship characterized by the law as ‘proximity’ or ‘neighbourhood’, and the situation must be one where it is fair, just and reasonable that the law should impose the duty of the given scope upon the one party for the benefit of the other. In the present case there is no difficulty over the first of the three headings. A solicitor must foresee that if he fails to prepare a will as instructed by his client, and arrange for it to be duly executed, the disappointed beneficiaries will suffer financial loss. After the client’s death nothing can be done to remedy the solicitor’s negligence. The estate will pass to those entitled under any valid unrevoked will and, subject thereto, to those entitled on an intestacy. In the nature of things it will then be too late for the solicitor to prepare a will. Furthermore the solicitor must foresee that the payment of damages to his deceased client’s estate would achieve nothing. This would not provide a means of recompensing the client for the solicitor’s professional negligence. Any money paid to the estate would simply pass under the very testamentary dispositions which fail to carry out the testator’s instructions. Money paid to the estate would not go to the disappointed beneficiaries.

A special relationship

I turn next to consider whether there is between a solicitor and intended beneficiary a relationship of proximity and whether it is fair, just and reasonable that there should be a liability imposed on the solicitor to compensate the intended beneficiary. I shall consider these two headings together, because there is no real demarcation line between them. They shade into each other. Both involve value judgments. Under the third heading the court makes its assessment of the requirements of fairness, justice and reasonableness. Likewise, although less obviously, built into the concept of proximity or neighbourhood is an assessment by the court that in a given relationship there ‘ought’ to be liability for negligence. These two headings are no more than two labels under which the court examines the pros and cons of imposing liability in negligence in a particular type of case. This is well illustrated in the instant case where some of the points which fall for consideration could happily be considered under either heading.

Mr Matheson submitted the authorities establish that, in cases of pure economic loss, there is no liability for damages in negligence unless there is actual and foreseeable reliance by the plaintiff on the defendants. Reliance is a prerequisite to the existence of the necessary close and direct relationship. I do not agree. As I read the speeches in the recent House of Lords decisions, their lordships have emphasised the importance of reliance in pure economic loss cases. But I do not find in those speeches any indication, still less express statement, that in all cases reliance is a prerequisite to the existence of liability. Indeed, the thrust of the speeches in Caparo v. Dickman and Murphy v. Brentwood is against any attempt to lay down hard-and-fast boundary lines.

I can take, as one example of a case where there was no reliance, the decision of the Court of Appeal in Ministry of Housing and Local Government v. Sharp [1970] 2 QB 223. There the local authority’s officer negligently issued a clear official search certificate to a would-be buyer of land. The buyer completed his purchase on the strength of the certificate. In consequence the Ministry lost the benefit of a charge for repayment of compensation previously registered by it in the local land charges register. There was no question of the Ministry relying on anything done by the local authority or the officer. Nevertheless the court held the local authority liable in damages to the Ministry for the negligence of its official. I see no difficulty with this decision. A buyer is entitled to rely on the accuracy of a search certificate. A duty of care is owed to him. He has a remedy against those concerned if he suffers loss by an error in a carelessly prepared certificate. It would be remarkable if, when the local authority’s duties are carried out negligently, a buyer has a remedy but an incumbrancer does not.

Nor is there an analogy between the shareholders and future investors in Caparo v. Dickman and intended beneficiaries in the instant case. In the Caparo case the House of Lords decided there was no reason in policy or in principle why the necessary close relationship should be held to exist either between the auditors and future investors or between the auditors and existing shareholders in respect of future purchases of shares. The statutory duties of auditors were not imposed by Parliament for the protection of investors in the market, and as a buyer of further shares an existing shareholder stands in no different position from any other investing member of the public: see Lord Bridge of Harwich at pages 623D–F and 627E–F, Lord Oliver at pages 642F–H, 649H–650C and 653E–G and Lord Jauncey of Tullichettle at pages 662A–D. The present case is quite different. The very purpose of the employment of the solicitor is to carry out the client’s wish to confer a particular testamentary benefit on the intended beneficiary. There is no other purpose. If the solicitor negligently fails to achieve that purpose, justice requires there should be some remedy available.

A further point taken by Mr Matheson was that since the testator is under no duty to confer a gratuitous benefit on his intended beneficiaries, there is no reason to impose on the testator’s professional adviser a duty which goes further than that on the testator. A testator does not owe a duty of care to his beneficiaries. If he fails to make a will, or if he makes a will but carelessly fails to have it properly witnessed, the disappointed beneficiaries cannot sue the testator for damages for breach of a duty of care. So, runs the argument, why should the disappointed beneficiaries be able to sue the testator’s solicitor when they cannot sue the testator himself, however careless he may have been?

In my view there is here a feature of fundamental importance in this case. In general, and always leaving reliance cases on one side, a solicitor owes a professional duty of care to his client and no one else. He is subject to professional rules and standards, and he owes duties to the court as one of its officers. But within that framework it is to his client alone that he owes a duty to exercise the standard of skill and care appropriate to his status as a solicitor. Thus, in general, when acting for a seller of land a solicitor does not himself owe a duty of care to the buyer: see Gran Gelato Limited v. Richcliff (Group) Ltd [1992] Ch. 560. In the ordinary course of adversarial litigation a solicitor does not owe a duty of care to his client’s opponent: see Al-Kandari v. J. R. Brown & Co [1988] QB 665, per Lord Donaldson of Lymington MR at p. 672 and per Bingham LJ at p. 675. Further, and a little closer to the present case, when advising a client about a proposed dealing with his property in his lifetime a solicitor does not owe a duty of care to a prospective beneficiary under the client’s will who may be prejudiced by the dealing: see Clarke v. Bruce Lance & Co [1988] 1 WLR 881. I observe, in passing, that the lifetime dealing there did not have as its object the benefit of the plaintiff to whom the property had been specifically given under an existing will.

So one asks oneself: is the position different regarding instructions for the preparation and execution of a will? if so, why? After all, such instructions are no more than one particular type of instructions given by clients to a solicitor. Why should a solicitor be liable to a third party in such a case but not in others? These are pertinent questions. Before attempting to answer them I must mention another point also urged by Mr Matheson. He submitted it would not be fair or reasonable to impose liability on the solicitor in favour of the intended beneficiary, because that would be to give the beneficiary the benefit of a claim for professional negligence for services under a contract not made by him and in respect of which he has made no payment.

I think it must be frankly recognised that if the court holds a solicitor liable to an intended beneficiary, what the court is doing is fashioning an effective remedy for the solicitor’s breach of his professional duty to his client. I do not shrink from this. If this sounds heretical to some, I pray in aid the observation of Deane J. in the High Court of Australia in Hawkins v. Clayton (1988) 62 ALJ 240, 259, in a case concerned with a solicitor’s liability to make reasonable efforts to locate the executor named in a will held by him in safe custody when he learns of the testator’s death:

‘The law of contract and the law of tort are, in a modern context, properly to be seen as but two of a number of imprecise divisions, for the purpose of classification, of a general body of rules constituting one coherent system of law.’

Here, a coherent system of law demands there should be an effective remedy against the solicitor. The law of contract is unable to provide the remedy. In some cases, where the purpose of a contract is to confer a benefit on a third party, the purpose can be achieved in the event of breach, by the court making an order compelling the party in breach specifically to perform his obligation to make a payment or confer some other benefit on a third party: see Beswick v. Beswick [1968] AC 58. That route is not available here. The solicitor did not agree to confer a benefit on the intended beneficiary. He agreed to take steps to enable his client to do so. Specific performance of that agreement is no longer possible once the client has died. I have, indeed, considered whether a remedy for breach of contract could be shaped whereby, the client having lost the opportunity to make a gift to the intended beneficiary, (1) his estate should be regarded as having lost a sum equal to the amount of the intended gift, and (2) the executors should hold that sum, when recovered from the solicitor, upon trust for the intended beneficiary.

In the end I can see no need for the law to grapple with the difficulties raised by this and other possibilities when there is to hand a simple remedy in negligence. Instructions to prepare a will are different from other instructions to a solicitor. The failure to carry them out properly results in the client’s purpose being thwarted but leaves the client’s estate with no effective remedy. There is good reason why the solicitor should be liable for a third party in this very special situation. If the mistake, such as faulty witnessing, is discovered in the client’s lifetime and corrected, the third party will suffer no loss. The cost of preparing and signing a new will would be recoverable by the client from the solicitor. But it is different if the client dies at a time when, because of a breach of the solicitor’s professional duty, the will for whose preparation the solicitor was responsible has not been duly prepared and signed. Then it is eminently fair, just and reasonable that the solicitor should be liable in damages to the intended beneficiary. Otherwise there is no sanction in respect of the solicitor’s breach in his professional duty. Thus there is a special relationship between the solicitor and intended beneficiary which should attract a liability if the solicitor is negligent.

It is true that the effect of holding the solicitor liable to the intended beneficiary will be to enable him to take advantage of the professional duty owed by the solicitor to the client. But if the solicitor was not so liable, he would go Scot free. He could commit a breach of his professional duty with impunity. That cannot be right. That cannot accord with the objectives the law seeks to attain when imposing upon solicitors and other professional advisers a duty to exercise due professional skill and care. The resources of the law must be sufficient to fill what otherwise would be a serious lacuna. As Bingham LJ observed in Simaan General Contracting Co v. Pilkington Glass Ltd (no. 2) [1988] 1 QB 758, 782:

‘Just as equity remedied the inadequacies of the Common law, so has the law of torts filled gaps left by other causes of action where the interests of justice so required’.

Other objections

Mr Matheson raised several further points which, he submitted, showed that the liability question cannot be answered in favour of the intended beneficiaries without giving rise to unacceptable distortions in the law. He contended the anomalies would be such that the remedy would be capricious in operation. He submitted, first, that to impose liability on the solicitor would be effectively to double the size of the client’s estate. This is incorrect. The damages are payable to the disappointed intended beneficiary, not to the deceased’s estate. Those entitled to the deceased’s estate receive a windfall in the sense that the deceased did not intend the estate should go to them. But that does not assist the solicitor’s case. That is the direct and foreseeable consequence of the solicitor’s breach of his duty of his client. Because of his negligence the client’s money did not reach the right pockets. The law is requiring him to put that right in the only way it can be done.

A further objection concerned the effect of an agreement between the solicitor and client releasing the solicitor from liability for negligence or limiting the extent of his liability. It was said that if a duty of care to an intended beneficiary exists, this duty is independent of the solicitor’s duty to his client. Accordingly an agreement of this nature between solicitor and client would not affect the solicitor’s duty to the intended beneficiary. This would be an unfair result. I agree that unless the duty to the beneficiary is correspondingly limited, the result would be unfair. However, I can see no overriding reason why the duty should not be limited in this way. Nor did Robert Goff LJ in Leigh and Sillavan v. Aliakmon Shipping Co. Ltd. [1985] 1 QB 350, at 397, where he outlined a principle of transferred loss. This approach did not altogether find favour with Lord Brandon of Oakbrook when the case reached the House of Lords ([1986] AC 785, 819–20). So this is an important subject which would call for careful examination if, which seems unlikely, it were ever to arise in practice in this field. It can be safely left until then.

Another objection concerned conflicting interests of potential beneficiaries. There is no difficulty here. The preparation by a solicitor of a second will in favour of different beneficiaries is not a breach of any duty owed by the same solicitor to the prospective beneficiaries under an earlier will also prepared by him. The duty on the solicitor to prospective beneficiaries is to exercise professional skill and care in carrying out the testator’s instructions, giving him all necessary and appropriate advice. The duty goes no further than that. When the first will has been duly prepared and executed, the solicitor has fully discharged his duties and obligations to the beneficiaries under that will. He owes no obligation to them which would inhibit him from subsequently accepting and carrying out instructions from his client to draw up a new will with different legacies.

Once it is appreciated that the solicitor’s duty goes as far as I have shortly stated but not further, I do not see any scope for difficulties over conflicts of interest. The making of a new will and its terms may well affect the existing potential beneficiaries differently: the proposed changes may be advantageous to one, prejudicial to another and neutral so far as a third is concerned. The solicitor’s duty throughout remains as I have stated. The fact that this duty is also owed to an intended beneficiary does not alter its nature.

Mr Matheson raised the spectre of the solicitor being liable to an indeterminate class. The intended beneficiaries might include persons unborn at the date of the testator’s death; and their interests might not vest finally for many years. I do not agree. Liability is not to an indeterminate class. Liability is to the particular beneficiary or beneficiaries whom the client intended to benefit through the particular will. If a case of negligence should ever occur where an intended beneficiary was not born when the client died, for example, the intended beneficiaries might embrace all the testator’s grandchildren whenever born, any difficulties this may cause can then be faced and resolved. This factor, of possible difficulties in imagined and unusual cases, cannot weigh heavily in the scales when set against the ordinary case in which the intended beneficiaries are a small number of identified people. The development of the law should be principled and coherent, but there must be room for special cases inwhich some loose ends, or even jagged edges, may be inevitable.

In this regard it is sensible to remember that Ross v. Caunters was decided in 1979. In practice the decision does not seem to have given rise to problems. Furthermore, the decision has generally been welcomed, or at least not made the subject of criticism, by the leading commentators: see, for example, Atiyah, An Introduction to the Law of Contract, 4th edn., p. 395; Fleming, The Law of Torts, 7th edn., p. 167; Salmond & Heuston on the Law of Torts, 20th edn., pp. 215, 217; Treitel, The Law of Contract, 8th edn., p. 539; and Winfield and Jolowicz on Tort, 13th edn., pp. 88, 96, 106.

Finally, it was said there are insurmountable difficulties over periods of limitation. If an intended beneficiary has a cause of action against a negligent solicitor, time would run either from the date when the will was made or ought to have been made or from the date of death. The first of these possibilities would be unsatisfactory; a beneficiary intended to take under a will can hardly be said to suffer loss during the testator’s lifetime. The second would be unsatisfactory, because it would mean that time might not begin to run indefinitely, subject only to the 15 year longstop under section 14B of the Limitation Act 1980. Again these are points best left to be resolved in a case in which they call for decision. If there are problems here, this would not be the first instance of difficulties over limitation in the field of negligence. Even if there are difficulties, as to which I express no view, this does not militate significantly against the existence of a liability in negligence in this type of case.

In my view Ross v. Caunters is still good law. I venture to echo Sir Robert Megarry’s prophecy (at p. 321) that wherever the bounds of negligence become finally drawn, they will be wide enough to give an affirmative answer to the liability question.

The solicitor’s liability to his client

For completeness I mention that some argument was addressed to us in support of the proposition that a solicitor’s liability to his client lies only in contract; he is not also liable in tort. It was submitted that the decision of Oliver J. to the contrary in Midland Bank Trust Co. Ltd v. Hett, Stubbs & Kemp [1979] Ch. 384 was wrong. The decision of the Court of Appeal in Groom v. Crocker [1939] 1 KB 194 has never been overruled (see Lee v. Thompson (1989) 40 EG 89), and it derives more recent support from the approach of Lord Scarman in Tai Hing Cotton Mill Ltd v. Liu Chong Hing Bank Ltd [1986] AC 80, 107. It is not necessary to pursue this point. The conclusion on the liability question must be the same whether the solicitor is liable to his client in contract alone or in both contract and tort.

Inter vivos gifts

I have confined my observations to intended testamentary benefits. In Ross v. Caunters Sir Robert Megarry (at pp. 322–3) summarised his conclusions in terms equally applicable to a case where the solicitor is instructed to carry through a transaction of gift to a third party during the client’s lifetime. I see no reason to doubt that a similar principle applies to such case, because the same difficulty arises after the client’s death. However, that is not a matter I need pursue on this appeal.

Breach of duty

During the hearing of the appeal the solicitors applied for leave to adduce further evidence with a view to seeking a retrial. We refused that application, for the reason set out in the judgment of Farquharson LJ.

As already mentioned, the Judge concluded that Mr Jones and his firm were in breach of their professional duty in not exercising due diligence in the preparation of Mr Barratt’s will. A month elapsed from 17 July, when Mr Jones received the written instructions he had asked for, until he left the office for his holiday. Thereafter a fifth week passed before Mr Barratt went away on holiday. There was no evidence before the Judge on the firm’s usual practice when carrying out instructions to prepare a will. Mr Jones’s explanation for the delay in preparation of what was, after all, a very simple will was that when he received the letter of instructions he was suspicious and did not believe the letter truly set out Mr Barratt’s wishes. When he received the letter he telephoned Mr Barratt who, in effect, confirmed this and said that he still wanted to leave something to his former son-in-law Peter Gould. According to Mr Jones, he told his client he was shortly to go on holiday; he was heavily committed and would be unable to see Mr Barratt until after his return.

The Judge rejected this explanation. This is hardly surprising, having regard to the terms of Mr Jones’s internal memorandum of the 16 August. There is nothing in the memorandum to suggest that Mr Jones had any thought that Mr Barratt might have been subject to undue influence. The Judge concluded that if the reason for inaction was not idleness, it was concern on Mr Jones’s part that his (Mr Jones’s) friend Peter Gould would end up with nothing if the will were changed. He acquitted Mr Jones of deliberate dishonesty in his evidence, although his conduct at the time invited censure.

Given these findings I can see no ground which would justify this court interfering with the Judge’s overall conclusion that there was a breach of professional duty by the defendant firm in carrying out Mr Barratt’s instructions. The letter of instructions was incomplete to the extent that it did not set out who were to be the executors, nor did it contain residuary gifts. But I think the Judge was entitled to conclude that Mr Jones effectually convicted himself out of his own mouth. By feeling obliged to put forward specious justifications for his delay, Mr Jones recognised implicitly that Mr Barratt had not received the professional service to which he was reasonably entitled.


I can see no reason to doubt that if Mr Jones and the probate department had got on with the matter as they should have done a new will, under which each plaintiff would have benefited at least to the extent of the £9,000, would have been signed by Mr Barratt before he went on holiday. The firm already had in its files the details of Mr Barratt’s estate obtained for the March 1986 will. The respects in which the new instructions were incomplete were largely formal, and a brief meeting at most would have sufficed to deal with them. Mr Barratt was anxious to put right without delay what he perceived as a wrong done by him to his daughters.

In expressing the contrary view the Judge relied heavily on the evidence of Mrs Ducros, who used to live at 86 Whitecroft Road, as demonstrating there was an element of uncertainty about Mr Barratt’s precise intentions. I am unable to agree there was any relevant uncertainty. Mrs Ducros’ evidence was that at Maxine’s wedding on 16 August Mr Barratt wanted to unburden himself. He told her that ‘the girls’, meaning his two daughters, would be alright, that the house would be sold and divided between them, and that the grandchildren would have about £200 each from the building society. This, of course, does not tally neatly with the letter of instructions. But the factor common to both is that the girls were to be far and away the major beneficiaries. The discrepancies do not suggest either that Mr Barratt was in such a state of indecision he might not have made a new will at all or that under his new will the daughters might not have received at least the sum of £9,000, whether by way of pecuniary legacies or as residuary legatees or as specific legatees of the house. On this point the Judge was drawing an inference from primary facts found by him. I have to part company from him on the proper inference.

Accordingly I would allow this appeal, set aside the Judge’s order, and enter judgment for each plaintiff in the sum of £9,000.

Notes to Case 25

1. Lawyers can through their negligence harm the interests of: (i) their clients and (ii) third parties. Increased litigation under the first heading, spurned no doubt by liability insurance becoming compulsory (as it is in Germany or France) or widespread (as it is in England and the USA), has led in some countries (e.g. the USA) to the flourishing of courses on Professional Responsibility. England, on the other hand, still refuses to follow suit, steadfastly clinging to archaic immunities accorded to barristers (but not solicitors who, in fact, have to be insured) for work done in court (see Rondel v. Worsley [1969] 1 AC 191). In those cases, where liability is possible in theory (but by no means easily engaged in practice), two problems have attracted the attention of the courts. The first is connected with the difficulties of assessing what would the chances have been of the disgruntled client/plaintiff but for the negligence of his legal adviser. The second is connected with the curious, if not unjustified, willingness of some American courts to set the standard of expected care at a very low level (note, for example, Lucas v. Hamm 56 Cal. 2d 583, 364 P. 2d 685 (1961) where the court was prepared to characterize as non-negligent an attorney who failed to take proper notice of the rule against perpetuities on the ground that it is widely misunderstood by the profession!). (For a comparative study of the problems of professional malpractice (including medical malpractice which is treated with less indulgence by the courts) see chapter 6 of vol. XI of the International Encyclopedia of Comparative Law.)

Case 15 deals with the second type of situation and it invites comparison with such cases as Biakanja v. Irving 49 Cal. 2d 647, 320 P. 2d 16 (1958) (pleaded in tort), Lucas v. Hamm, above, (pleaded in contract and tort) and Heyer v. Flaig (74 Cal. Rptr. 225; 443 P. 2d 161 (1969)) decided in tort in the USA, and Ross v. Caunters [1980] Ch. 297 (pleaded in tort only) in England. Similarities and differences must again be stressed.

2. The first point to note is, again, the tendency of German law to base its solution on contract. That contract cannot easily explain all such cases is obvious not only by the court’s own review of the precedents but also by a certain unease it experiences in utilizing the notion of contract with protective effects vis-à-vis third parties or the theory of ‘transferred loss’ (Drittschadensliquidation), both briefly explained above (Ch. 2, section A, A. 2. (d) (iii) (e)). The first German decision on the subject—BGH NJW 1965, 1955; JZ 1966, 141 note Lorenz—relied on the notion of contract with protective effects vis-à-vis third parties and found for the plaintiff, though one is inclined to believe that the theory of transferred loss might have been an even more appropriate device. (This case is discussed in English by Lorenz in ‘Some Thoughts about Contract and Tort’ Essays in Memory of Professor F. H. Lawson (1986) 86 ff.) Incidentally, the granting of an action against the attorney (for a sum equal to the void legacy) can in practice result in a ‘double-legacy’—a point that has troubled some commentators. (See Kegel, ‘Die “lachenden Doppelerben”: Erbfolge beim Versagen von Urkundspersonen’, Festschrift für Flume (1978) I, 545. This argument, however, has been rejected both in England (White v. Jones, supra) and in Germany (see: BGH 12 June 1979, Zeitschrift für das Gesamte Familienrecht 1980, 133 and Lorenz and Markesinis (1993) 56 MLR 558 et seq.)

3. The contract versus tort approach must not only be noted but also examined, especially since some American decisions (e.g. Heyer v. Flaig 74 Cal. Rptr. 225; 443 P. 2d 161 (1969)) have gone as far as suggesting that the distinction may be meaningless. It is submitted that it is not, and American law may be the poorer for ignoring some of the practical consequences of a proper doctrinal characterization. Here are some:

(a) Jurisdiction. The choice-orientated approach adopted by conflicts rules in cases of contract seems more appropriate in these types of cases than the locus of the accident tort approach. In any event, it would be odd if the dispute between plaintiff and defendant/debtor were subject to one rule and the relationship between defendant/debtor and creditor were governed by a different law. (See Fig. 2.) We can, of course, again allow the contractual relationship to influence the tort action (as Tobriner J. suggested in Heyer v. Flaig 70 Cal. 2d 223, 449 P. 2d 161 (1969)); but why not opt more openly for the more straightforward contractual reasoning? (The potential impact that the choice of contract or tort can have on jurisdiction can be seen in factual variations of cases such as Bryant Electric v. City of Fredericksburg 762 F. 2d 1192 (1985) and Blake Construction Co. v. Alley 353 SE 2d 726 (1987) both of which represent variations on the J’Aire and Junior Books theme.)

(b) Liability for omissions. Traditional tort theory would deny liability where the debtor/defendant has remained inactive rather than acted badly. A contractual solution could make the defendant/debtor liable both for non-feasance and misfeasance. Should this be regarded as excessively onerous for the defendant/debtor it may still be avoided through the use of Drittschadensliquidation or the German notion of contract with protective effects vis-à-vis third parties, since this, it will be remembered, makes the debtor liable only for the bad or delayed performance of the secondary obligations of the contract. (The first German case was one of omission: BGH NJW 1965, 1955; JZ 1966, 141 (with note by Lorenz). Cf. the factually almost identical cases of Gartside v. Sheffield Young and Ellis [1983] NZLR 37 and Hof Amsterdam, NJ 1985, 40—the last two resorting to tort doctrine.)

(c) Measure of damages. The tort and contract measure of damages probably remains different despite Lord Denning’s attempt to propose a compromise in H. Parsons (Livestock) Ltd. v. Ottley Ingham and Co. Ltd. ([1978] QB 791). A contractual solution would favour the award of full expectation damages to the plaintiff/third party and this, again, is supported by German law. In the USA a tort solution would further justify a claim for punitive damages. Such a ridiculous claim was, in fact, made in Heyer v. Flaig.

(d) The limitation period. Both the length and starting point of the limitation period differ in contract and tort. It is submitted it would be impracticable if one relationship (plaintiff–defendant/debtor) were subject to one rule and the other relationship (debtor–creditor) governed by a different rule. This appears to be in principle the correct starting point but it must be admitted that in some cases it may cause problems. Thus, in the frustrated beneficiary type of case the right of the beneficiary to sue the testator lawyer is not identical to that of the testator since the harm suffered by each is different. Also, if the contractual period of limitation were to apply in exclusion of all tort remedy, it could harm the beneficiaries’ interests in those systems where the contractual period of limitation is not substantially longer than the tort period of limitation.

(e) The standard of care. Few lawyers in England or in the USA seem to have considered points (a)–(d). More concern, however, has been shown about the standard of care that would be appropriate to the tort action and whether it would be different from that found in the debtor–creditor (contractual) relationship. German law, proceeding on contract theory, has not experienced this difficulty; and German writers, considering a hypothetical (for them) tort action have had no difficulty in saying that the tort standard of care should be determined by the contract. (Schlechtriem, Lorenz, op. cit.) American courts have taken a similar view; and both Lord Roskill (in Junior Books) and Robert Goff LJ (in The Aliakmon) saw little difficulty in adopting this reasoning in English law. Of course, the contractual solution avoids at a stroke these real or imaginary difficulties.

(f) Exemption clauses/defences. It has never been doubted in German law that the contractual debtor can oppose against the third party/plaintiff all defences, etc., he may have against the contractual creditor. The contractual solution makes this answer indisputable. In an obiter dictum in The Aliakmon Robert Goff LJ suggested that something similar should happen where the action was founded in tort ([1985] QB 350 at pp. 397–8) but details have not been worked out. Strangely enough, at least one first instance judgment in England has considered a variant to this problem.

In Southern Water Authority v. Carey ([1985] 2 All ER 1077) the court was not asked to decide whether the defendant sub-contractor (debtor) could oppose against the owner/plaintiff (third party) the exemption clauses that he (the sub-contractor) had in his contract with the main contractor (creditor); instead the court was faced with a claim by the sub-contractor to oppose against the owner a clause which the main contractor had in his contract with the owner. The court thought he could do so and, I believe, the solution displays a misunderstanding of how this triangular relationship should work. It is submitted that had a contractual explanation been used this misunderstanding would not have occurred.

In Carey the plaintiffs (or, rather, their predecessors in title) engaged the first defendants as consulting engineers and the second defendants as main contractors in the building of certain sewage works. The second defendants in turn engaged the third and fourth defendants as sub-contractors. Clause 30 of the work contract between the plaintiffs and the second defendants (main contractors) contained an exemption clause. Subsection (VI) of this clause stated that ‘for the purposes of this sub-clause the Contractor contracts on his own behalf and on the behalf of and as trustee for his sub-contractors, servants and agents’. The main contractors having ceased trading, the action for defective workmanship and materials proceeded against the two sub-contractors. At this stage it is worth noting that while the fourth defendant/sub-contractor had in his contract with the main contractor a term almost identical to Clause 30 of the main contract, the third defendants/sub-contractors, when negotiating with the main contractor, had insisted on their own, different, limitation clause. (Though we are nowhere told in what respect the two sets of terms differed, one must assume that the contractor’s terms were more advantageous to the defendants/sub-contractors than their own terms, for otherwise why did they try to invoke their protection?) The Court, applying the tort reasoning of Junior Books, accepted that the plaintiff had a cause of action. Unlike Junior Books, however, the plaintiff was faced with a defence put forward by the defendant. This was based on an exclusion clause contained not in his (the sub-contractor’s) contract with the main contractor (in which case the position could have been similar to The Aliakmon), but on the exclusion clause contained in the contract between the plaintiff and the main contractor (and which the third defendant/sub-contractor had, as stated, rejected). Diagrammatically the difference from The Aliakmon case can be shown as in Fig. 2.

As stated, the court was prepared to allow a tort action by the employer against the sub-contractor but was also willing to give the latter the protection of the clause contained in the contract between employer and main contractor. One reason was that the nature of this limitation clause was relevant in defining the scope of the duty of the sub-contractor in tort. Another explanation was the use of the notion of volenti. The judge expressly excluded the latter concept (ibid. at p. 1085a); but later in his judgment he appears to have come very close to accepting it (or something very similar to this concept) since he said: ‘As the plaintiffs . . . did . . . choose to limit the scope of the sub-contractor’s liability, I see no reason why such limitation should not be honoured.’ (Ibid. at pp. 1086 f.) Is the reasoning of this decision and its result acceptable? I confess I have doubts as to both.

It is first of all interesting to note that in Carey the court did not follow The Eurymedon (New Zealand Shipping Co. Ltd.) v. A. M. Satterthwaite and Co., The Eurymedon [1975] AC 154) reasoning in order to achieve its result. The artificiality of the contractual reasoning in that case has been noted by many lawyers; but it was reaffirmed by the Privy Council in Port Jackson Stevedoring Pty. Ltd., The New York Star v. Salmond and Spraggon (Australia) Pty. Ltd. ([1981] 1 WLR 138) so the judge in Carey was obviously anxious not to challenge it openly. On the other hand, it was not used either, so those opposed to The Eurymedon result and reasoning might be able to argue that in future it may be limited to the special shipping context of such cases. This still, however, leaves open the question of how to explain Carey.

One could, de lege ferenda, construe Carey as a case of contract in favour of third party (and, for brevity’s sake I include in this term the contract with protective effects vis-à-vis third parties). After all, Carey is, in one respect, very similar to Junior Books; the owner could be seen as the third party beneficiary of the contract between sub-contractor (promisor) and main contractor (promisee). This construction, of course, is unacceptable de lege lata given the English notion of privity. But even if it were accepted it would still not explain the second part of Carey namely the allowing to the promisor the defences that the promisee had against the third party. For in contracts in favour of third parties it is the reverse that is allowed: namely the promisor can oppose against the third party the defences he has in his contract with the promisee.

Can one then justify the outcome through tort reasoning? The judge in Carey did so by allowing the tort action to be shaped by the contract to which the plaintiff was a party but the defendant was not. In this he was clearly influenced by certain dicta contained in Lord Roskill’s judgment in Junior Books v. Veitchi. (Lord Roskill refers there to ‘a relevant exclusion clause in the main contract’.) With respect I do not find Lord Roskill’s dicta as obvious as that, my inclination being to interpret them as referring to the second contract (sub-contractor–main contractor) and not to the first contract (between contractor and owner). I incline in favour of this view largely because Lord Roskill seemed to rely heavily in his relevant passage on Hedley Byrne; and there, apart from the fact that there were no contractual relations stricto sensu, the exemption clause was inserted by the second bank/defendant when replying to the plaintiff’s bank and not by the plaintiff’s bank (which was merely transmitting the information) to the plaintiff. The clause, in other words, was in the second not the first relationship. My own understanding of the Roskill dictum is also shared by the Court of Appeal in Muirhead v. Industrial Tank Specialities Ltd. ([1986] QB 507; Reynolds in (1986) 10 LMCLQ 97, 106 probably agrees). On the other hand, it must be admitted that this is not how the judge understood it in Carey; and his interpretation also has the important support of Professor Treitel (in (1986) 10 LMCLQ 294, 302). Clearly, therefore, the dictum has led to some confusion and should be clarified at the earliest possible opportunity. Nevertheless, even if it were eventually to be classified in the way it was understood in Carey it still does not, in my opinion, justify the actual outcome in that case. For in that case we were in fact faced with an interesting variant to the problem we have been discussing. Because in Carey the judge was faced with two sets of exemption clauses, one in the contract between owner and main contractor and one in the contract between main contractor and sub-contractor; and he allowed the sub-contractor the benefit of the first clause even though at the time of conclusion of the relevant contracts the sub-contractor had rejected it in favour of his own clause. I see no commercial reason justifying this result; nor do I think we can say that the plaintiff assented to a limited liability when the defendant, at the time of the making of the contracts, rejected the relevant limitation clause.

4. Contract or tort, the new action must be kept under control, especially where third parties/non-clients are concerned. American decisions are legion—see, for example: Victor v. Goldman 74 Misc. 2d 685, 344 NYS 2d 672 (1973); Pelham v. Griesheimer 93 Ill. App. 3d 751, 417 NE 2d 882 (1981), aff’d. 92 Ill. 440 NE 2d 96 (1982); Guy v. Liederbach 459 A. 2d 744, 501 Pa. 47 (1983); Needham v. Hamilton 459 A. 2d 1060 (App. DC 1983); Flaherty v. Weinberg 303 Md. 116, 492 A. 2d 618 (1985); Walker v. Lawson 514 NE 2d 629 (Ind. App. 1987); Stinson v. Brand 738 SW 2d 186 (Tenn., 1987); Mentzer and Rhey Inc. v. Ferrari 367 Pa. Super. 123, 532 A. 2d 484 (1987); Hale v. Groce 304 Or. 281, 744 P. 2d 1289 (1987); Simon v. Zipperstein 32 Ohio St. 3d 74, 512 NE 636 (1987). Schreiner v. Scoville 410 N.W. 2d, 679 (Iowa: 1987); Licatae v. Spector, 225 A. 2d, 28 (1966: Connecticut) 61 ALR 4th 63–501 suggests that a majority of jurisdictions now favour plaintiffs—at least in the Ross v. Caunters type of situation. The end result, however, is not all that different from the more restrictive European (German and English) approach. Nevertheless the question remains whether the multi-criteria test enunciated in Biakanja v. Irving is more workable and (on balance) fairer than the contractual third party beneficiary doctrine. In this context consider: (i) why the six factors mentioned in Biakanja (320 P. 2d 16, 19) were subsequently reduced to five (the requirement that ‘moral blame [be] attached to the defendant’s conduct’ has been dropped); (ii) whether the various criteria are cumulative or whether one of them is particularly important; (iii) that equal uncertainty seems to prevail on the requirements needed to be satisfied for the invocation of the third party beneficiary doctrine. In particular, who must intend to confer a benefit on the third party: the promisee (better view) or the promisee and the promisor? From the vast literature see: Bauman, ‘Damages for Legal Malpractice’, 61 Temple L. Rev. 1127 (1988); idem, ‘Lawyer Liability to Non-Clients’, 97 Dickinson L. Rev. 267 (1993); Boston, ‘Liability of Attorneys to Non-Clients in Michigan: a re-examination of Fridman v. Dozore and a rule of limited liability’, 68 Univ. of Detroit L. Rev. 307 (1991); Dugdale, ‘Solicitors’ Liability to Third Parties’ [1984] NZLJ 316 ff.; Eisenberg, ‘Attorneys’ Negligence and Third Parties’ 57 NYUL Rev. 126 (1982); Gross, ‘Contractual Limitations on Attorney Malpractice Liability: An Economic Approach’ 75 Kentucky LJ 793 (1986–7); Hazard, ‘Lawyer Liability in Third Party Situations: The Meaning of the Kaye Scholer Case’, 26 Akron L. Rev. 395 (1993); Hilliker, ‘Attorney Liability to Third Parties: A Look to the Future’ 36 Depaul L. Rev. 41 (1986); Keeton, ‘Professional Malpractice’ 17 Washburn LJ 445 (1978); Markesinis, ‘Fixing Acceptable Boundaries to the Liability of Solicitors’ 103 (1987) LQR 346; idem, ‘Doctrinal Clarity in Tort Litigation: A Comparative Lawyer’s Viewpoint’, 25 The International Lawyer 953 (1991); Morley, ‘Privity as a Bar to Recovery in Negligent Will-Preparation Cases’ 57 Univ. of Cincinnati L. Rev. 1123 (1989); Peterson, ‘Extending Legal Malpractice Liability to Non-clients’ 61 Wash. L. Rev. 761 (1986); Probert and Hendricks, ‘Lawyer Malpractice: Duty Relationships Beyond Contract’ 55 The Notre Dame Lawyer 708 (1980); Rolph, ‘Solicitors’ Liability to Non-clients in Negligence’, 15 The Advocates’ Quarterly, 129 (1993); Rubenstein, ‘Attorney Malpractice in California: The Liability of Who Drafts an Imprecise Contract or Will’ 24 UCLA Law Review 422 (1976); Walker, ‘Attorney Liability to Third Parties for Malpractice: The Growing Acceptance of Liability in the Absence of Privity’ 21 Washburn LJ 48 (1981–2). Gates, ‘Lawyers’ Malpractice’ 37 Mercer L. Rev. (1986) gives interesting statistical information.

8. ‘Other rights’ (family relationships)

Back to top

This page last updated Thursday, 01-Dec-2005 11:04:42 CST. Copyright 2007. All rights reserved.