Robinson v Tait [2001] NZCA 217; [2002] 2 NZLR 30; (2001) 9 NZCLC 262,651 (18 June 2001)

Last Updated: 13 December 2011

IN THE COURT OF APPEAL OF NEW ZEALAND

CA 70/00

	BETWEEN	A W ROBINSON AND OTHERS
		Appellant

	AND	W D TAIT AND OTHERS
		Respondent

Hearing:	5 April 2001
Coram:	Thomas J Keith J Blanchard J Tipping J McGrath J
Appearances:	F Miller and B H Fletcher for Appellants R C Laurenson and S J Peacock for Respondents
Judgment:	18 June 2001

JUDGMENTS OF THE COURT

Judgments
	Para Nos.
Thomas J	[1]-[27]
Keith, Blanchard, McGrath JJ	[28]-[87]
Tipping J	[88]-[99]

THOMAS J

Thompson’s case

[1] I have had the instruction of reading the draft judgment of Blanchard J and agree that under s 37(6) of the Securities Act 1978, as a matter of statutory interpretation, each of the directors are severally liable to repay the subscriptions in issue. My interest in writing separately, however, relates to the decision of the High Court of Australia in Thompson v Australian Capital Television Pty Ltd [1996] HCA 38; (1996) 186 CLR 574.
[2] Mr Miller, who appeared for the appellant, was forthright in his criticism of this decision. He submitted that the High Court’s reasoning was illogical and indefensible. I take the opposite view. I do not believe that the High Court’s reasoning can be faulted.

The judgments of the High Court of Australia

[3] The judgment of Brennan CJ and Dawson and Toohey JJ, begins (at 581) with an affirmation of the common law. Where there was a joint tort there could be only one action and one judgment for the whole amount of damages to which the plaintiff was entitled. That is to say, the cause of action was one and indivisible so that when judgment was obtained on it, whether against one or more of the joint tortfeasors, the cause of action merged in the judgment and precluded further recovery against any remaining tortfeasors. The learned Judges observed that this was also the basis of the rule that the release of one joint tortfeasor released all the others.
[4] The Judges then referred to s 11(2) of the Law Reform (Miscellaneous Provisions) Act 1955 (ACT), the equivalent provision to s 17(1)(a) of the Law Reform Act 1936. They held (at 582-583) that “the effect” of s 11(2) is that the cause of action against joint tortfeasors is no longer one and indivisible. Reference was made to the decision of the Privy Council in Wah Tat Bank Ltd v Chan [1975] AC 507, and the English Court of Appeal’s decision in Bryanston Finance Ltd v de Vries [1975] QB 703, both of which held that the common law doctrine relating to the merger of the cause of action against all joint tortfeasors in the first judgment recovered against any of them had been abolished in its entirety. These decisions, of course, related to the judgment rule and were, therefore, not directly in point. Rather, they were used to verify the initial step in the Judges’ reasoning; that is, that the statute had abolished the unity rule in its entirety in respect of successive judgments. It followed, they held (at 584), that a cause of action against joint tortfeasors is no longer one and indivisible. The dictum of Brennan J in XL Petroleum (NSW) Pty Ltd v Caltex Oil (Australia) Pty Ltd [1985] HCA 12; (1985) 155 CLR 448, at 459 was cited with approval. In that case, Brennan J said (at 466) that the effect of the statutory provision upon the joint and several liability of joint tortfeasors was far-reaching in that the unity of the cause of action against all joint tortfeasors was severed by the provision. The resulting implications were not merely procedural.
[5] The learned Judges in Thompson then summarised their reasoning (at 584):

Similarly, under s 11(2) of the ACT Act it is no longer the case that the victim of a tort committed by joint tortfeasors has only one cause of action; the cause of action is no longer one and indivisible. The concept of a single wrong and a single cause of action having gone, the rule that a release given by one joint tortfeasor releases any others must have gone with it, for that rule is nothing more than another aspect of the same thing, namely, that there is only one cause of action against all joint tortfeasors in respect of the one tort. In other words, once the cause of action is by statute no longer one and indivisible, there is no conceptual basis for the rule that the release of one joint tortfeasor releases the others. The rule must therefore be taken to have been impliedly abolished by the statute. (Emphasis added).

[6] The Judges concluded (at 584) that the absence of an express statutory provision providing that the release of one joint tortfeasor does not release the other joint tortfeasors in respect of the same tort did not affect the necessary implication that the common law rule in relation to releases must have gone at the same time as the cause of action became no longer one and indivisible.
[7] Gummow J, with whom Gaudron J agreed, also held that s 11 operates so that the unity of the cause of action against all joint tortfeasors is severed. Gummow J began (at 608) with the proposition that a release given to, and an accord and satisfaction effected with, one of a number of parties jointly or jointly and severally liable, contractually discharges the others. The learned Judge undertook a comprehensive survey of the case law. He held (at 611) that it is against this background that, in modern times, the rationale for the discharge of the remaining joint tortfeasors by the release of one of them should be seen as supplied, not by satisfaction, but solely by the ground advanced in Duck v Mayeu [1892] 2 QB 511, at 513, that is, that the one and indivisible cause of action having been released, all persons otherwise liable are also released. It follows, Gummow J said, that if s 11 of the Law Reform Act operates upon a footing which denies the inseverable unity of the cause of action against joint tortfeasors, then the substratum upon which the authorities rest as to discharge as a consequence of release is removed. (See also the learned Judge’s reiteration of the point at 613-614).

Parliament’s intent

[8] With respect, it is a mistake to approach the issue as a question of statutory interpretation and seek to ascertain whether or not Parliament intended to “abolish” the release rule. To do so, and triumphantly proclaim that Parliament had no such discernible intent is to miss the point. Parliament had no such intent and, for that reason, the language of “abolition” is inappropriate. Rather, it is to be accepted that Parliament did not seek to address the release rule, although it (or its advisers and draftspersons) would have been fully aware of that rule. But there is nothing untoward or definitive in that admission. As Professor Beatson has recently pointed out in an excellent article, statutory reform tends to advance by statutes correcting particular defects in the law. He suggests that this process, described as “legislative microsurgery”, will make the law even more of an amalgam of common law and statute. See J Beatson, “The Role of Statute in the Development of Common Law Doctrine” (2001) 117 LQR 247, at 252. In the present case, and prior to the enactment of s 17, Parliament was concerned with a major problem, that is, the situation where a plaintiff who has sued one tortfeasor to judgment and recovered nothing from him or her cannot afterwards proceed against another person who is equally liable. Parliament moved to remedy that particular problem.

The removal of the one and indivisible cause of action

[9] In remedying the particular problem Parliament removed the basis of the common law principle that the cause of action against joint tortfeasors is a single unified cause of action. The effect or consequence of the statute, which admittedly relates solely to the judgment rule, is to remove the substratum necessary for the operation of the release rule. As Professor Fleming puts it, “by authorising successive actions, the unity of the common law action against all tortfeasors was severed in its entirety. See Fleming The Law of Torts (1998 – 9th ed) at 36. Or, as the author of Winfield and Jolowicz on Torts (1998 – 15th ed) at 732, observes, “the fact that even joint tortfeasors may be sued in successive actions has ‘heavily compromised’ the logic of unity of liability upon which it [the release rule] rests”.
[10] The same point may be made in another way. Where a plaintiff has obtained a judgment against one joint tortfeasor which is not satisfied in whole or in part, so that the plaintiff then sues another joint tortfeasor, the latter’s liability does not arise at that point. Liability extends back to the date when the cause of action arose. The original cause of action is still the foundation of the plaintiff’s claim and the liability of the second joint tortfeasor. In effect, several liability has existed from that date and it can no longer be suggested that the cause of action is one and indivisible.
[11] As the one and indivisible cause of action was also the basis of the release rule, what basis then exists for that rule? It would be incongruous and artificial to suggest that at one and the same time the joint tortfeasor’s liability to a plaintiff is severed from the date of the plaintiff’s cause of action for the purpose of the judgment rule, but remains a unified and indivisible cause of action should the plaintiff enter into a discharge of liability in respect of one joint tortfeasor only. The flaw in Mr Miller’s argument is his failure to recognise that the second (or third or fourth) joint tortfeasor who is sued following an unsatisfied judgment against the first joint tortfeasor, is sued on the basis of the same cause of action, and that his or her potential liability arises from the date of that cause of action. The following diagrams may assist to illustrate this point:

[12] Fig 3 shows the artificiality and incongruity of seeking to impose on the joint liability of joint tortfeasors a unified and indivisible liability for the purposes of the release rule. As a result of s 17, the necessary substratum has been summarily removed.

The doctrine of merger

[13] Formulations of the judgment rule refer to the cause of action merging in the judgment. This element is, of course, absent from the release rule. But it does not affect the essential point that the judgment rule, like the release rule, was based on the notion of a one and indivisible cause of action. It was because the cause of action was one and indivisible that it merged in the first judgment obtained against the defendant.
[14] Joint liability is not necessary for merger to apply. The authors of Clerk and Lindsell on Torts (2000 – 18th ed) para 32-14, at [1693], point out that when an action is brought before a competent court and proceeds to final judgment, the original right of action is destroyed. Professor Street puts it this way: “The original cause of action is terminated by its merger in the judgment.” (Street on Torts (1999 - 10th Ed) at p 567). The principle is simply an aspect of the res judicata doctrine. Where joint or joint and several liability exists the principle is extended to the cause of action in respect of all parties so liable simply because it is one and indivisible. The doctrine of merger is a consequence of the concept of the one and indivisible cause of action just as the release rule is a consequence of that same unity. It does not vitiate the High Court’s reasoning in Thompson.

The demise of joint liability?

[15] Mr Miller submitted that for this Court to follow Thompson would bring about the demise of joint liability altogether. I cannot accede to this claim. In the first place, there is something odd about seeking to preserve the release rule in order, allegedly, to preserve the concept of joint liability, especially when Parliament has decided otherwise in respect of the judgment rule. More critically, in the second place, the abandonment of the release rule does not destroy the concept of joint liability at all. That concept remains a valid legal concept. What it does mean is that many of the law’s technical requirements relating to joint liability, such as the release rule itself, are rendered inappropriate to discharge that liability.

Brooks’ case

[16] Nor do I agree with Mr Miller that this Court in Brooks v New Zealand Guardian Trust Co Ltd [1994] 2 NZLR 134, fully confronted the reasoning which emerged later in Thompson. In Brooks’ case, this Court was principally concerned with the question whether s 17, as a matter of statutory interpretation, abrogated the release rule. Thus, the Court concluded (at 140) that, while it is clearly arguable that a logical corollary of the 1936 Act would be an abrogation of the rule, “the grounds are strong for holding that no such intention can be attributed to the legislature in the 1935 Act in England or the 1936 Act in New Zealand”. (Emphasis added). As explained above, it is accepted that Parliament had no such intent. Rather, the necessary consequence of Parliament’s intent to abrogate the judgment rule is to annul the common law’s one and indivisible rule, which was also the basis of the release rule. (I should add that discarding the release rule does not necessarily mean that the Court in Brooks would have been required to reach a different conclusion. The question whether the release was intended to discharge the directors as well as the company would still fall for determination as a matter of interpretation).

Procedure

[17] During the course of oral argument there was some discussion as to whether the phrase “joint and several liability” is used in a substantive sense or is directed to the remedy. Although it is fair to say that he did not press the point, Mr Miller submitted that the concept is a remedial or evidential concept in that it relieves the plaintiff of the burden of proving which part of a single injury was caused by which defendant. With respect, I find the distinction between substance and procedure in this context a barren distinction. As Brennan J said in XL Petroleum (NSW) Pty Ltd v Caltex Oil (Australia) Pty Ltd, supra, at 466) (see para [4] above), the effect of the statutory provision upon the joint and several liability of joint tortfeasors was far-reaching. The unity of the cause of action against all joint tortfeasors was severed by the provision and “the resulting implications are not merely procedural”.

Another route – following Parliament’s lead

[18] I do not think it matters greatly what route is chosen to dispense with the release rule. It can be accepted either that the logical corollary of s 17 is the dissolution of the unified and indivisible basis of the cause of action against joint tortfeasors, or that when applying the common law the Court should follow the lead set by Parliament. What is important is to realistically acknowledge, as Gummow J has so ably demonstrated, that the case law supporting the unified indivisible rationale for the release rule has been undermined and that there is no longer any rational basis for the rule. For the Courts to now persist with the rule is to be either pedantic or indifferent to the policy and direction which Parliament, the supreme law maker, has chosen to pursue.
[19] Even if unable to accept the reasoning of the High Court of Australia, therefore, this Court should not be hesitant in following Parliament’s lead. The release rule has been strongly criticised by respected jurists for many years. Professor Glanville Williams, for example, has asserted: “There can be no doubt that the rule relating to releases has worked badly, and that with its modern refinements it has become, as Justice Cardozo put it, an incumbrance and a snare”. See Glanville L Williams “Joint Obligations” (1949), para 64, at 137, quoting Cardozo Law and Literature (1931) at 51. Nothing short of unthinking allegiance to an exaggerated perception of the doctrines of precedent and stare decisis can now counter this sustained criticism.
[20] Moreover, the law reports contain many examples of decisions where the Courts have been openly influenced by the policy and direction of particular legislative measures. Many of these cases were referred to by this Court in R v Moke and Lawrence [1996] 1 NZLR 263, at 268-270, in support of the principle that the Courts should, in developing the common law, proceed in parallel and by analogy with the course pursued by Parliament. See also R v H [1997] 1 NZLR 673, at 695. It will be sufficient, however, to refer to the timely article by Professor Beatson, supra.
[21] Professor Beatson soundly refutes the “oil and water” approach to statute and common law (confirming in the course of doing so what every judge in the realist tradition would accept happens in practice). He refers to numerous examples of the analogical use of statute. The application by equity of statues of limitation to situations to which they do not apply is perhaps the most striking example (at 252). But the learned Professor refers (at 252-255) to many other examples, and to judicial dicta, to the effect that developments in the common law ought to proceed upon a parallel rather than a diverging course in the area left to it by statute. Professor Beatson also points out that statute law has been of importance in determining the scope and operation of common law doctrines based on public policy and may be of vital assistance in determining the ambit of policy. Provided that the courts are sensitive to the scope of the statute, the statutory context may be of vital assistance in determining the ambit and relevance of policy (at 256).
[22] It is therefore a mistake to assume that, because a statutory provision does not overtly govern a situation, it is irrelevant to the determination of an issue. Parliamentary supremacy must mean that Parliament is to be accepted as the ultimate source of legal policy and the direction of the law. The notion that, because a statute may only deal with part of the subject matter before the Court, that part not dealt with is not “the law” is archaic today. Such a negative conclusion does not follow. Parliament may have had no intent on the point or may have intended to leave further development to the common law. The case where Parliament enacts a comprehensive statutory regime and the Courts must make a pragmatic decision as to whether or not Parliament intended to eject the common law from that particular field is entirely different. Clearly, in abrogating the one and indivisible basis for the purpose of the judgment rule, Parliament did not assume the whole field. But that does not mean it is to be attributed with a positive intent to preserve the release rule. Rather, that rule remains to be applied by the Courts in accordance with the lead which Parliament has given in relation to the one and indivisible rule.

The law is left in an unsatisfactory state

[23] Whatever route is taken, therefore, the Court should be prepared to discard the release rule. Not to do so, but to hold that a release does not discharge the joint tortfeasors where the liability is “joint and several”, as distinct from “joint”, is to leave the law in an anomalous and unsatisfactory state. A few instances will suffice to make this point.
[24] First, the liability of joint tortfeasors is effectively joint and several liability where judgment is concerned but not joint and several where a settlement prior to judgment is involved. Take, for example, the position of three joint tortfeasors A, B, and C. The plaintiff obtains judgment against A, but that judgment is only satisfied in part. As permitted by s 17 the plaintiff then sues B. B executes a release. Under the release rule the plaintiff cannot pursue C, but had the plaintiff pursued B to judgment he or she would, by virtue of s 17, be able to pursue C to judgment. The inherent disharmony is self-evident.
[25] Secondly, as the plaintiff will not be bound by the release rule whenever the liability is joint and several, arguments that liability is purely “joint” and not “joint and several” will be the certain outcome. Such arguments are patently sterile.
[26] Finally, joint and several liability is likely to prevail in contract, simply because draftspersons will follow that time-honoured format. Yet, with concurrent liability now accepted in contract and tort the position will exist where a defendant may be jointly and severally liable in contract but only jointly liable in tort in respect of the same claim. A release will not discharge another person jointly and severally liable in contract but will discharge that same person in tort!

Conclusion

[27] For the above reasons, while I do not wish to depart from the majority in dismissing the appeal, I would prefer to adopt the rationale of the High Court of Australia in Thompson. In enacting s 17, Parliament has destroyed the concept that a victim of a tort committed by joint tortfeasors has only one cause of action and that cause of action is one and indivisible. As the release rule is only an aspect of the same concept, it must be accepted that, to use Gummow J’s phrase, the substratum for the rule has been removed. Alternatively, I would be prepared to abandon the release rule on the basis that the Court should follow Parliament’s lead and seek to develop the law in parallel with the statute.

KEITH, BLANCHARD, McGRATH JJ
(DELIVERED BY BLANCHARD J)

[28] Section 37 of the Securities Act 1978 forbids the making of any allotment of securities offered to the public for subscription unless, at the time of the subscription, there was a registered prospectus relating to the security. It further provides that, where subscriptions are received by or on behalf of an issuer but, for that reason, securities cannot be allotted, or for any reason are not allotted, the issuer must ensure that the subscriptions are kept in a trust account and are repaid to the subscribers, together with any interest earned thereon, as soon as reasonably practicable.
[29] This appeal is concerned with subs(6) of s37 which provides that if any such subscriptions are not so repaid within two months after their receipt by or on behalf of the issuer, then the issuer and all the directors thereof “shall be jointly and severally liable to repay the subscriptions”, together with interest at a stipulated rate from the date of their receipt by or on behalf of the issuer. But the subsection also states, in a proviso, that a director is not so liable “if he proves that the default in the repayment of the subscriptions was not due to any misconduct or negligence on his part”.
[30] The issuer in this case was Fortex Group Ltd which is insolvent. When Fortex collapsed in 1994, it was discovered that s37 applied to subscriptions paid by some of its employees for shares in deferred payment share schemes. Their subscriptions, amounting to $1,668,467, had not been held in a trust account. They were not repaid within the two month period.
[31] In 1996 a group of employees, acting as representatives, sued Fortex and the trustee of the schemes. They did not at that time sue the Fortex directors. On 18 December 1996 the employees entered into a deed of settlement with Fortex and the trustee. The detail of that settlement appears in a judgment of this Court in the present proceeding delivered on 18 December 2000 (the interim judgment). In summary, Fortex, through its receivers, agreed to pay $950,000 to the plaintiff employees and those they represented and it was agreed that, upon completion of payment, the claimant on whose behalf the payment was made should be deemed to have

waived and abandoned any claim which he or she may have against the Company or the Trustee in respect of any of the said share schemes and, without limitation to the foregoing, to have settled his or her claim in the Proceedings and to have agreed to their discontinuance.

[32] Subsequently, the present respondents, who were amongst those who participated in the settlement, have brought this representative proceeding against the directors of Fortex, in reliance on s37(6), claiming $598,913, which is said to be the balance of the unrefunded subscriptions.
[33] In the High Court six of the directors, the present appellants, sought summary judgment, arguing that, because the liability of the directors was joint as well as several, they have been released by the deed of settlement. They rely on the common law rule that a release of one joint obligor discharges the entire obligation and so releases any other joint obligor (Duck v Mayeu [1892] 2 QB 511). Master Venning took the view that, for the reasons he gave, which are summarised in this Court’s interim judgment, there had not been a release of anyone other than Fortex and the trustee.
[34] In the interim judgment it was determined that the settlement deed had operated as a release, unqualified by any reservation of the right to bring a claim against the directors. We said that there was an insufficient basis for any implication that Fortex and the trustee were agreeing that the employees should continue to enjoy the right to sue the directors, with the adverse consequence for Fortex and the trustee that the directors, if found liable, might then seek indemnity or contribution from them. But the Court directed that certain questions should be argued before a Court of five, as has now occurred, namely whether the common law rule survives in New Zealand and, if it does, whether it applies to a joint and several obligation or merely to a joint obligation. With the greater focus resulting from the arguments which have now been heard, it is clear that the question which will be determinative of the directors’ appeal is whether the release rule applies in the particular statutory setting of s37(6) where Parliament has expressly said that the directors are to be jointly and severally liable along with the issuer for the return of the subscriptions.

The release rule and the judgment rule

[35] Duck v Mayeu confirmed that

a release granted to one joint tortfeasor, or to one joint debtor, operates as a discharge of the other joint tortfeasor, or the other joint debtor, the reason being that the cause of action, which is one and indivisible, having been released, all persons otherwise liable thereto are consequently released. (p513)

[36] It was also settled law that, because a plaintiff’s cause of action merges into the judgment which the plaintiff obtains against the defendant, a judgment against one joint obligor entirely replaced the cause of action in all respects and thereby discharged any other joint obligors (Brinsmead v Harrison (1872) LR 6 CP 584).
[37] These two rules have been called, respectively, the release rule and the judgment rule.

Section 17 of the Law Reform Act 1936

[38] It has been held unanimously by the High Court of Australia in Thompson v Australian Capital Television Pty Ltd [1996] HCA 38; (1996) 186 CLR 574 that the common law rule, as stated in Duck v Mayeu, was abolished by a statutory provision substantially in the same terms as s17(1) of the Law Reform Act 1936 in this country. Accordingly the effect of that section was much debated in argument and it is convenient to set it out:

17 Proceedings against, and contribution between, joint and several tortfeasors

(1) Where damage is suffered by any person as a result of a tort (whether a crime or not)—

(a) Judgment recovered against any tortfeasor liable in respect of that damage shall not be a bar to an action against any other person who would, if sued, have been liable as a joint tortfeasor in respect of the same damage:

(b) If more than one action is brought in respect of that damage by or on behalf of the person by whom it was suffered, or for the benefit of the estate, or of the wife, husband, parent, or child of that person, against tortfeasors liable in respect of the damage (whether as joint tortfeasors or otherwise), the sums recoverable under the judgments given in those actions by way of damages shall not in the aggregate exceed the amount of the damages awarded by the judgment first given; and in any of those actions, other than that in which judgment is first given, the plaintiff shall not be entitled to costs unless the Court is of opinion that there was reasonable ground for bringing the action:
(c) Any tortfeasor liable in respect of that damage may recover contribution from any other tortfeasor who is, or would if sued in time have been, liable in respect of the same damage, whether as a joint tortfeasor or otherwise, so, however, that no person shall be entitled to recover contribution under this section from any person entitled to be indemnified by him in respect of the liability in respect of which the contribution is sought.

Argument for the appellants

[39] In an attractively presented argument, Mr Miller submitted that s17 did not abrogate the release rule, supporting the view taken earlier by this Court in Brooks v New Zealand Guardian Trust Co Ltd [1994] 2 NZLR 134. He said the breach of statute by Fortex was to be treated as a tortious breach (of statutory duty) by the directors; that s17 did not destroy the unity of the joint tort and that the judgment rule, which s17(1)(a) abolished, was a separate procedural rule from the release rule (also a rule consequential on the unity of the joint tort), which still remained. He argued that Thompson cannot be reconciled with the legislative history of the joint tort rule and that the reasoning of the High Court of Australia is inconsistent with the continued existence of joint torts. It was submitted that any further reforms should come from Parliament and that policy arguments exist in favour of retaining the release rule. The argument that it is a trap for the unwary was said to be overstated. If the rule is to be changed, counsel submitted, there would be a need for further rules paralleling those found in paras (b) and (c) of s17(1). Abandonment of the release rule was also said to have consequences for the doctrine of accord and satisfaction.
[40] Mr Miller submitted that joint liability is imposed so as to relieve the plaintiff from having to establish which part of the loss was caused by each wrongdoer, a question often likely to be found impossible to resolve. He observed that joint wrongdoers commit torts jointly. They are described as being jointly and severally liable or, as in the heading to s17, as joint and several tortfeasors, only because the plaintiff is entitled to sue each tortfeasor separately – each being liable for the whole of the damage suffered by the plaintiff.
[41] The several aspect of the liability is therefore remedial, not substantive. Counsel said that s17 did not abolish the joint nature of a joint tort; it merely dispensed with one awkward consequence namely that, because of the doctrine of merger, the right of action against the joint wrongdoers was replaced by the judgment against one of them, even if that judgment subsequently went unsatisfied. Conceptually, however, counsel said that the liability of the wrongdoers is still unified.
[42] Mr Miller cited The Koursk [1924] P 140 and In re EWA, a debtor [1901] 2 KB 642 in support of the view that the release rule applies to joint wrongs whether the liability is joint or joint and several. He said that the position is essentially the same when the joint obligation is contractual.
[43] Turning to s37(6), counsel said that there was only one omission to repay the subscriptions, namely that of Fortex. The statute prevents the directors from saying that they were simply acting on behalf of the company and are not personally liable. Counsel submitted that the statutory obligation is not contractual in nature, although he accepted that the moneys had been paid to the company pursuant to a contract, but not one to which the directors were parties. It was said that there was a closer analogy to the tort of breach of statutory duty. Here the statute provided the remedy which tort law would otherwise provide. In making directors severally as well as jointly liable the drafter was looking to procedure.
[44] Mr Miller pointed out that, as this Court had remarked in Brooks (at p140), in most cases the directors of a company are not mere agents but persons for whom the company is directly liable. Their actions are attributed to the company and treated as the very actions of the company itself.
[45] Counsel argued in support of retaining the release rule generally, because it avoids both multiple actions and any circuity of actions through claims for contribution. If it did not exist, he contended, settlements by a defendant would be discouraged as there could be no guarantee of finality. Counsel said that plaintiffs are normally advised on their claims and competent legal advice avoids the risks associated with the release rule. It operates only in default of a provision in a settlement. A settling defendant is entitled to know whether it will continue to be exposed to a claim for contribution by another defendant against whom the plaintiff subsequently proceeds. If a plaintiff intends to reserve the right to sue other joint obligors or joint tortfeasors, such reservation should be clearly stated in the deed of settlement.

Argument for the respondents

[46] Whereas the argument for the appellants approached the construction of s37(6) by reference to the position of joint tortfeasors, Mr Laurenson, for the respondents, concentrated on joint and several contractors, paying particular attention to certain quite recent English cases in which, it was argued, there had been a shift in view as compared with In re EWA. He characterised the obligation of Fortex and the directors as a statutory debt. He drew attention to the observation of Gummow J in Thompson (p604) that generally, in contrast to the position in tort, a judgment against one contractor who is jointly and severally liable does not bar the several remedy against the others. In a carefully researched submission counsel traced the development of the rule in contract, placing emphasis on the judgment of Parke B in King v Hoare [1844] EngR 1042; (1844) 13 M & W 494; 153 ER 206, which drew a distinction between joint contractors (release of one of whom will discharge the others) and joint and several contractors, where the several obligations were said to give different remedies to the obligee.
[47] Mr Laurenson suggested that later authorities, including EWA, had overlooked the distinction. He noted comment in Halsbury 4ed Vol 9(1), para 1089, fn.1 and in Chitty on Contracts, 28ed para 18-017 on the illogicality of applying the release rule in favour of someone who is severally liable for an obligation. He submitted that the matter has now been put on a correct basis in Johnson v Davies [1999] Ch 117, 127, where it was said that the release of the debt owed by B to A would in logic have no effect on the several debt owed to A by another co-debtor, C; that the relevant question is whether the release agreement between A and B precludes A from enforcing the debt owed by C, which, as Neill LJ pointed out in Watts v Aldington (1993), The Times, 16 December 1993; [1993] CA Transcript 1578, ought to be determined by the surrounding circumstances and taking into account, not only the express words used in the document, but also any terms which can properly be implied.
[48] Mr Laurenson submitted that Parliament had deliberately chosen to impose a several obligation upon Fortex and the directors as well as their joint obligation. He submitted that, in view of the purposes of the legislation – to restore the subscriptions, which were a liquidated sum – it cannot have been intended that a compromise with one of them should release the others from their separate liability in circumstances where the obligation has not been entirely satisfied by repayment of the whole of the subscriptions.

Liability in tort

[49] We are of the view that, in the end, the issue in this case comes down to the construction of s37(6) in its particular statutory context. But, as that construction is likely to be informed by the rules of common law concerning enforcement of joint and several obligations, which may have influenced the choice of statutory language, we proceed to consider those rules by way of background. We begin with the rules concerning joint tortfeasors.
[50] Joint tortfeasors are persons who together commit a single wrong against the victim. They are to be contrasted with concurrent tortfeasors who separately commit different wrongs against the same victim causing the same or overlapping injury. This judgment is not concerned with concurrent tortfeasors, to whom the release rule does not apply because, although the loss may be the same, the obligations arising from the injuries inflicted are not: see Allison v KPMG Peat Marwick [2000] 1 NZLR 560 for an extended discussion of the position of concurrent wrongdoers, particularly at para [139].
[51] The position of joint wrongdoers had, until the High Court of Australia’s decision in Thompson, appeared relatively straightforward. Persons who had together committed a single tort against the plaintiff were jointly responsible for it, each being liable for the whole of the damage (Cocke v Jennor [1792] EngR 516; (1614) Hob 66; 80 ER 214). Only one sum of damages can be awarded for the joint tort (Sir John Heydon’s Case [1572] EngR 356; (1612) 11 Co Rep 5a; 77 ER 1150). This approach avoided the need for the plaintiff to prove which part of the loss had been caused by the acts and omissions of each of the individual tortfeasors. The tort was regarded as unitary – one wrong committed in whole by all tortfeasors - and not able to be subdivided. But, because each was liable for the whole, there were both joint and several remedies. Take a case where there were two joint wrongdoers. The plaintiff could sue both together and obtain one judgment against them both; or the plaintiff could sue one of them separately and obtain a judgment for the full amount of the loss against that one alone. Since, however, there was only one tortious obligation, any step which involved a release of the plaintiff’s right against either tortfeasor was regarded as necessarily bringing to an end the rights against the other. Those steps might involve a settlement which discharged the settling tortfeasor, usually upon payment of an amount in or towards the plaintiff’s loss, or the obtaining of a judgment against one tortfeasor, whether sued separately or along with the other wrongdoer.
[52] The reason why the obtaining of a judgment was regarded as bringing to an end the plaintiff’s right to sue the wrongdoers for the unitary tort was that the right merged in the judgment. Henceforth, the right to enforce the judgment replaced the right to sue in tort. Because of the unity of the tort, if the plaintiff’s right was replaced by a judgment against one tortfeasor, it could no longer exist as the basis for an action against the other. In Brinsmead v Harrison the Court cited the judgment in Brown v Wootten (1604) Cro Jac 73; 79 ER 62:

[A]ll the Court held the plea to be good; for, the cause of action being against divers, for which damages uncertain are recoverable, and the plaintiff having judgment against one person for damages certain, that which was uncertain before is reduced in rem judicatam, and to certainty; which takes away the action against the others...

[53] But the judgment rule was very inconvenient. A well-advised plaintiff could avoid releasing one wrongdoer pursuant to a compromise until the settlement moneys were actually paid over or secured; or the plaintiff could withhold a release but, in exchange for a payment, agree not to further pursue his or her rights against the settling tortfeasor. (As there was, at common law, no right of one tortfeasor to claim contribution against another, the settling defendant was not exposed to any further liability (Merryweather v Nixan [1799] EngR 669; (1799) 8 TR 186; 101 ER 1337)).
[54] But it was different when judgment was obtained. Merger into the judgment was automatic. The plaintiff could not reserve the right to sue the judgment debtor, let alone another joint tortfeasor, even if the judgment sum were not able to be recovered. The plaintiff lost the right to sue regardless of whether there was any recovery under the judgment.
[55] From the judgment debtor’s perspective, there was also no possibility of its claiming contribution if the judgment were wholly or partially satisfied. It was actually this no-contribution rule which was the stimulus for the reforms in England in 1935. The Third Interim Report of the Law Revision Committee recommended that a person adjudged to be liable to make any payment in respect of an actionable wrong should be able to recover contribution from any other person liable in respect of the same wrong or who, if sued separately, would have been so liable. Almost incidentally, in a single paragraph (para 11), the Committee commented that, in altering the law on this subject, it would seem desirable to alter the rule that the tort is merged in the judgment even though there is no satisfaction. Section 6 of the Law Reform (Married Women and Tortfeasors) Act 1935 in England and s17 of the 1936 Act in New Zealand were enacted to give effect to these recommendations.
[56] Section 17(1)(c) enables a tortfeasor to recover contribution from any other tortfeasor liable in respect of the same damage “whether as a joint tortfeasor or otherwise [concurrent tortfeasor]”. Para(a) provides that judgment recovered against a tortfeasor does not bar an action by the plaintiff against another joint tortfeasor. Para(b) contains two consequential reforms. The joint liability is capped at the sum awarded by the first judgment against any of the tortfeasors. It was thought necessary to have this cap in part because juries in separate proceedings might make inconsistent awards. As the Law Commission in England noted in 1977 in its Report on Contribution in the Law of Contract (at para 41), it is doubtful whether the cap can be justified now that jury trials have ceased to be the normal method of trying civil actions. The other reason for the cap was to prevent multiplicity of actions. However, as the Commission also pointed out, this is sufficiently done by the second of the consequential reforms in para(b), namely the stipulation that in a subsequent action the plaintiff is not entitled to costs unless the Court is of opinion that there were reasonable grounds for bringing the second action.
[57] In England it has been very much the predominant view – one accepted in the Law Commission report just referred to – that the Law Reform Act removed the bar created by the judgment rule but had no effect on the rule that a settlement which released one joint tortfeasor without reservation of the right to sue another joint tortfeasor also discharged the liability of that other. (The reform had, however, exposed a settling joint tortfeasor to a contribution claim where the settlement did contain a reservation of the right to sue another tortfeasor.)
[58] Lord Diplock expressed the prevailing view in a dictum in Bryanston Finance Ltd v de Vries [1975] QB 703, 732:

The technical common law doctrine of “release” is unaffected by the Act. Such an agreement still has the effect in law of releasing all other joint tortfeasors as well, though courts nowadays are reluctant to construe an agreement with one tortfeasor as a release rather than a covenant not to sue him, unless it is plain that the agreement was intended by the plaintiff to operate also as a release of the other joint tortfeasors from their liability.

[59] In the same case, Lord Denning was of the contrary opinion that, after settling with one tortfeasor, the plaintiff could sue another who was jointly liable but could recover only if awarded more than the settlement amount. This Court in Brooks noted these divergent opinions, but was not convinced that what it called the established release rule had been indirectly abrogated by statute.
[60] In Thompson, the joint judgment of Brennan CJ, Dawson J and Toohey J provided little analysis beyond saying that the effect of a provision similar to s17(1)(a) was that the cause of action against joint tortfeasors is no longer one and indivisible. They relied to a considerable extent on the decision of the Privy Council in Wah Tat Bank Ltd v Chan [1975] AC 507 where the Board held that para(a) of the equivalent of s17(1) “expresses with reasonable clarity the intention of the legislature to abolish the common law rule in its entirety”. With respect to the Judges in the High Court of Australia, however, it seems doubtful that the Board was speaking of more than the judgment merger rule, which it described as “a highly technical and unsatisfactory rule” (p515). The point in issue in Wah Tat was whether the reforming provision applied where two or more tortfeasors were sued in the one action or, as had been argued, only where they were sued in separate actions. The Judicial Committee was rejecting the view that the judgment rule continued to apply if the defendants were sued in the one action, which would have meant that the provision had created a ludicrous procedural distinction. The subject of release by means of a settlement is simply not mentioned.
[61] Gummow J, with whom Gaudron J concurred, traced the history of the common law rule and discussed its rationale in some detail. He came to the same conclusion, namely that the basis of the release rule, irrespective of any policy arguments in its favour, is the indivisibility of the cause of action. Gummow J considered that the common law could no longer operate on that foundation in Australia because the statute had severed the unity of the cause of action. This conclusion is contrary to that of this Court in Brooks, which was not cited in Thompson.
[62] Because we are concerned in this case with the construction of a statutory liability, it is unnecessary for us to determine whether the High Court of Australia was right to take the view that the release rule, as it relates to joint tortfeasors, has been impliedly abrogated by the reform legislation (in New Zealand s17(1)(a)). The correctness of that view would seem to depend upon whether the abrogation of the judgment rule has the result that a plaintiff must now be considered to have ab initio separate causes of action against each joint tortfeasor – that they have each committed separate torts – but, presumably, still without need to prove a particular individual contribution to the overall loss; or whether, on the other hand, all that has happened is that the legislature has said that, despite the unitary nature of the tort, the consequence of entering judgment against one wrongdoer is not to be that a proceeding against another is barred. On the latter view, the judgment rule and the release rule are subsets of the unitary tort rule, with one of the subsets (the judgment rule) being reformed, without change either to the basic rule or to its other subset (the release rule). This may perhaps have been the view of Neill LJ in Watts v Aldington, for, at p10 of the unreported judgment, he commented that s3 of the Civil Liability (Contribution) Act 1978, which in England replaced s6 of the 1935 Act and extended the reform to contract as well as tort, was in terms limited to preserving a cause of action against the other tortfeasor “and does not sever a joint judgment”.
[63] Before leaving this subject, we make some brief comment about the policy justifications put forward on behalf of the release rule by counsel for the appellants. It is said to be unfair to a defendant that, where the plaintiff does not reserve the right to sue another tortfeasor, the settling defendant should be exposed to the risk of a contribution claim; and that the release rule avoids multiplicity and circuity of actions. Against this view it can be said, and we think with some strength, that any trap for the unwary should not be set for the plaintiff, who has been the victim of wrongdoing; that the law should favour the injured party by placing responsibility on the defendant (Allison at para [143]). Plaintiffs should be permitted to sue successively all those who have jointly caused injury unless they agreed not to do so.
[64] A settling defendant is exposed to a contribution claim only if the settlement is overly generous, so that the settling defendant has not met a fair share of the loss. This will be unusual where the defendant is worth (further) powder and shot. As to multiplicity of actions, the answer to argument in favour of the release rule would seem to be that the plaintiff must give credit, in any subsequent action against another tortfeasor, for the recovery under the settlement and is therefore at risk concerning costs if failing to obtain additional damages. For the reasons given by the Law Commission in England in 1977 (see para [56] above), a costs rule to this effect would seem to suffice.

Liability in contract

[65] Liability in tort is imposed by law upon the joint wrongdoers, but the liability of persons who jointly accept contractual obligations flows from the terms of the contract itself. Contracting parties who are together to undertake obligations may agree to do so jointly but not severally, or they may agree to do so both jointly and also separately (or severally). In the latter case there are multiple covenants – a joint obligation and a series of several obligations owed separately by each contractor to the party or parties on the other side of the contract. The rules concerning the effect of judgment and release could be expected to differ according to the nature of the obligation of each defendant.
[66] In New Zealand there has not been statutory reform of the relevant common law contract rules, as there has been in the United Kingdom by s3 of the Civil Liability (Contribution) Act 1978, which gave effect to the Law Commission’s Report in 1977. In New Zealand s17(1)(a) reforms tort law only.
[67] Where the obligation is joint but not several and the plaintiff elects to sue one only of the joint obligors, that obligor is usually entitled to insist upon joining any other joint obligors as co-defendants and may apply for a stay unless this has been done. (The exceptions, which are of no relevance in this case, relate to a joint party who is an undischarged bankrupt, outside the jurisdiction, a minor, a person protected by the limitation period, a member of a firm of common carriers, an undisclosed sleeping partner or an active partner of one who represented himself as being the sole contracting party (Halsbury, Vol 9(1), para 1080)). The right of a joint contracting party to insist on joinder of all other persons with the same liability was formerly in England invoked by a plea in abatement (Bristow v James [1797] EngR 437; (1797) 7 TR 257; 101 ER 962 in which the difference between the positions of joint contractors and joint tortfeasors was remarked upon). Subject to the foregoing exceptions, it is still the case that if the sued contractor does not seek a stay and the plaintiff proceeds to judgment, any other joint contractors are released. (In former times the others could then make what was known as a plea in bar.) This result occurs for the same reason as joint tortfeasors were similarly released before the enactment of s17(1), namely the unitary nature of the obligation.
[68] The classic exposition of the position of joint contractors is to be found in the judgment of Parke B in King v Hoare, which concerned an action in debt for goods sold and delivered by the defendant jointly with another seller. Parke B said:

We do not think that the case of a joint contract can, in this respect, be distinguished from a joint tort. There is but one cause of action in each case. The party injured may sue all the joint tortfeasors or contractors, or he may sue one, subject to the right of pleading in abatement in the one case, and not in the other; but, for the purpose of this decision, they stand on the same footing. Whether the action is brought against one or two, it is for the same cause of action. (p505; 210)

[69] This is still the law for joint contractors – for a recent example see Morris v Wentworth-Stanley [1999] 2 WLR 470, in which it was held that an accord and satisfaction with one debtor in respect of a joint debt discharged the other joint debtors unless there was an express or implied agreement that the creditor’s rights against them were preserved.
[70] But it appeared from King v Hoare that the position was not the same if the defendants had accepted both a joint and a several obligation. This view had already been stated in 1807 when Sir James Mansfield CJ, with the other Judges concurring, said in Ayrey v Davenport [1807] EngR 313; (1807) 2 Bos & Pul (NR) 474; 127 ER 714, concerning an action by the holder of a joint and several promissory note given by the defendant and Jeremiah Barber

If an action be brought against Barber, and he suffer judgment by default or confession, how can that discharge an action against the Defendant? The Plaintiff has a right to sue every one of the makers of this note to judgment; though it is true that he cannot have satisfaction more than once. And whether the judgment against Barber were obtained by cognovit [confession] or in any other way can make no difference. (p476; 715)

[71] The distinction was noted by Parke B in King v Hoare, as he made plain in the passage which follows immediately after the quotation from that case given above:

The distinction between the case of a joint and several contract is very clear. It is argued that each party to a joint contract is severally liable, and so he is in one sense, that if sued severally, and he does not plead in abatement, he is liable to pay the entire debt; but he is not severally liable in the same sense as he is on a joint and several bond, which instrument, though on one piece of parchment or paper, in effect comprises the joint bond of all, and the several bonds of each of the obligors, and gives different remedies to the obligee. (p505; 210)

[72] The distinction was noted by Earl Cairns LC in Kendall v Hamilton (1879) 4 App Cas 504 where the House of Lords held that partnership debts were joint debts. At p518 Earl Cairns commented:

If, therefore, this case is to be looked at as a case in which judgment has been recovered for a partnership debt against two out of three co-partners, it appears to me that, on the principle of King v Hoare, the judgment would be a bar at law to a subsequent action against the third co-partner; and I know of no principle on which a Court of Equity could hold the debt to be several for the purpose of preventing such a result. [Emphasis added]

[73] But it seems that later English courts chose to follow the rejection of the distinction in an even earlier case, Cheetham v Ward [1797] EngR 350; (1797) 1 Bos & Pul 630; 126 ER 1102. Cheetham was followed, without analysis, in a line of cases including Nicholson v Revill [1836] EngR 384; (1836) 4 Ad & El 675; 111 ER 941 and North v Wakefield (1849) 13 QB 536 and, most notably, in In re EWA a debtor [1901] 2 KB 642. (The views of Sir James Mansfield CJ and Parke B just referred to were overlooked or disregarded.)
[74] In re EWA concerned a joint and several judgment for a joint and several obligation under a guarantee. Collins LJ said, without expressly referring to authorities cited in argument, that it was “too late now to question the law – that where the obligation is joint and several, the release of one of two joint debtors has the effect of releasing the other” (p648). Judge Paul Baker QC followed this loyally in Deanplan v Mahmoud [1993] 1 Ch 151. However, Halsbury regards this approach as both “curious” (para 1089, fn1) and “surprising” (para 1090, fn3). Chitty (para 18-017) finds the supposed rule “illogical”, as did Professor Glanville Williams in his work on Joint Obligations (1949) at p135.
[75] Recently, however, there has been a correction. Johnson v Davies is the most modern English authority containing an extended discussion of the release rule in contract, although it was actually concerned with a joint obligation, not one which was joint and several. Considerable reliance was placed by the English Court of Appeal upon its decision in Watts v Aldington, which had involved a settlement with one of two persons against whom the plaintiff had obtained a joint judgment for defamation. (An action in debt on a judgment is no different in principle from an action in debt against joint contractors (Cocks v Brewer [1843] EngR 206; (1843) 11 M & W 51; 152 ER 711.)) The Court in Watts had emphasised that what mattered was not whether the settlement was a release or a covenant not to sue, but whether, in either event, it expressly or by implication reserved the plaintiff’s rights to enforce the judgment against the other judgment debtor. Commenting upon Watts, the Court in Johnson v Davies, speaking through Chadwick LJ, said that the liability of the judgment debtors had been plainly several as well as joint. Chadwick LJ continued:

In such a case, for the reasons explained in the judgments in this court, the relevant question is not whether the agreement between the creditor, A, and one of the co-debtors, B, releases the debt which B owes to A. Even if it did, that would, in logic, have no effect on the several debt owed to A by the other co-debtor, C. The relevant question is whether the agreement between A and B precludes A from enforcing the debt owed by C. It is in B’s interest that the agreement should have that effect – because, if it does not, C will be in a position (if he pays the debt which he owes to A) to seek contribution from B. It is in A’s interest that the agreement should not have that effect – because, prima facie, A will wish to recover from C the balance of the indebtedness. Given the opposing interests of A and B, the question is what have they agreed. As Neill LJ pointed out [in Watts], that has to be determined ‘having regard to the surrounding circumstances and taking into account not only the express words used in the document but also any terms which can properly be implied. ([1999] Ch 117, 127) [Emphasis in original]

[76] It is notable that the view was taken that any release of the debt owed by one of the obligors would not logically have the effect of releasing the several debt owed by the other obligor. As we have attempted to show, this view is in accordance with a line of 19th Century authorities. We consider that it is to be preferred to the other line which culminated in In re EWA.
[77] It may be understandable that an entirely unitary contractual obligation should be discharged by the release of one obligor, for it is a single obligation. There may be nothing remaining for the other obligor to perform. But where the contract also expressly creates several obligations performable by each obligor separately, which was presumably done for the very purpose of avoiding the characteristic of a unitary obligation just mentioned, it is hard to see why, unless a contrary intention is manifested, a release of one obligor should do more than releasing the unitary obligation of all, plus the separate obligation of that person. The separate obligation of the other co-obligor(s) is thus left extant. If the obligor who is being released does not wish the creditor to be able to continue to pursue the others, the settlement can of course provide accordingly, perhaps by stipulating that the right to pursue the co-obligor(s) is not reserved or, more directly, by expressly releasing the co-obligor(s).

The obligation under s37(6)

[78] This background is helpful in ascertaining the way in which the statutory obligations of Fortex and each director should be analysed.
[79] However, the determinative issue in this case is whether the release rule applies in the particular statutory setting of s37(6):

(6) If any subscriptions to which this section applies are not so repaid within 2 months after the date on which the subscriptions were received by or on behalf of the issuer (or, in any case to which subsection (2) of this section applies, within 5 months after the date of the date of the registered prospectus), the issuer and all the directors thereof shall be jointly and severally liable to repay the subscriptions, together with interest at the rate of 10 percent per annum from the date on which the subscriptions were received by or on behalf of the issuer:

Provided that a director shall not be so liable if he proves that the default in the repayment of the subscriptions was not due to any misconduct or negligence on his part.

(Note: The reference to 10 percent per annum was replaced in 1997 by reference to a rate prescribed from time to time by regulations.)

[80] The background demonstrates that the proper inquiry is as to what form of liability has actually been created by the sub-section. Tort law imposes only a joint or unitary obligation. Contractual obligations of co-obligors can be created either as a joint obligation only, or as a combination of the joint obligation and a separate individual (or several) obligation of each co-debtor, with the different consequences observed above.
[81] A company which fails to repay subscriptions in accordance with s37 is in breach of a statutory obligation. That is ordinarily characterised as a tort. Here, however, the section also states the remedy for the breach and, in imposing liability upon the directors, does so in language which is usually to be found in contracts where there is an intention that the liability of covenantors is to be separate as well as joint. The Court is concerned with the interpretation of an additional liability imposed for the purpose of better protecting the subscribers. Obviously Parliament was intending to widen the net and catch the directors when the company was unable to repay the subscriptions.
[82] Section 37(6) imposes personal liability on directors where the company fails to repay subscriptions within two months, unless they can show, in each individual case, an absence of misconduct or negligence. This Court said in its interim judgment that the fact that a particular defendant might succeed in a separate defence under the proviso would not operate as a severance of the joint aspect of the liability imposed by the legislation. It would simply be held that, because of the factual defence, that defendant was not one of those who were jointly and severally liable. But the proviso is, we think, an indication that there is an importance given to the existence of separate liability.
[83] The drafter has chosen to state the liability as being joint and several, just as a person drawing up a contract would do if it were desired to have separate obligations as well as a joint obligation amongst the contractors. Mr Miller urged the Court to find that, in referring to several liability, the statute merely invoked the procedural advantage enjoyed by the victim of a tort, namely the ability to sue joint tortfeasors separately. Counsel suggested that this was probably done only as a matter of caution; even without the words “and severally”, a plaintiff subscriber could have elected to sue the directors separately.
[84] We consider, however, that in subs(6) the words “jointly and severally” were used substantively to create a liability of that nature, rather than creating merely a joint liability enforceable both jointly and severally. Parliament was creating a form of liability that would otherwise not have existed. It could be expected to have described the several liability as a mere matter of procedure if no more than that had been intended. It is an unlikely specification. Can it really have been intended, for example, that a settlement with one director will release the issuer and the other directors, in the absence of full satisfaction? We think not.
[85] We read the words literally, as indeed they would now be read in a contract. The issuer and the directors are jointly liable, but each of them is also severally liable to make the repayment of the subscriptions. Because there is a separate obligation imposed on each director for the full amount, the subscribers can sue each director separately. A settlement with the company or with any of the directors may discharge the joint obligation and that person’s separate obligation but, subject to the terms of the settlement, would leave intact the subscribers’ right to pursue the separate obligations of each of the others.
[86] The settlement with Fortex “waived and abandoned” the claim against Fortex and the trustee. It contained no reservation of any right to bring a claim against the directors. It therefore did not preserve, for what it might in practical terms be worth, the joint obligation. But nothing in the document dealt in express terms with the separate obligation of the directors. This was not the subject matter of the settlement. For the reasons we have given, we consider that these obligations were additional to the joint obligation and are capable of continuing after the release of the joint obligation. Fortex and the trustee do not appear to have given any thought to the possibility of a claim against the directors and should not be taken to have intended to extinguish their separate obligations or to prevent the employees from pursuing the directors. Nor should the employees be taken to have intended to accept the $950,000 in full satisfaction of the subscriptions. They were compromising with the company and the trustee for less than the full face value of their claim.
[87] The Court being unanimous, the appeal is dismissed with costs of $8,000 to the respondents. The appellants must also pay the reasonable disbursements of the respondents, to be fixed if necessary by the Registrar.

TIPPING J

[88] I agree with the judgment prepared by Blanchard J. I shall not seek to replicate or emulate his comprehensive discussion of the development of this branch of the law. His Honour’s judgment clearly illustrates a distinction which I have always understood to apply between joint liability on the one hand, and joint and several liability on the other.
[89] In tort those liable jointly may be proceeded against severally, but their substantive obligation remains a joint one. There has never been any such thing in tort as a liability which is substantively both joint and several. Concurrent liability is several only. As to these points see The Law of Torts in New Zealand by Todd and Others (3rd ed: 2001) at 24.2.1ff, Cashfield House Ltd v David & Heather Sinclair Ltd [1995] 1 NZLR 452, and the exceptional joint and several liability of partners for torts committed by the firm created by ss13 and 15 of the Partnership Act 1908: compare s12, dealing with debts and other obligations, where the liability is stated to be joint only, in accordance with Kendall v Hamilton (1879) 4 App Cas 504. The references in s15 (and in s12 as regards the estate of a deceased partner) can hardly have been to several liability in only a procedural sense: see Lindley and Banks on Partnership (17th ed: 1995) at 13-12 (page 383).
[90] Obligations in tort arise by operation of law. This contrasts with the position in contract where the obligations which arise have been created by the voluntary acts of the parties. Thus contracting parties may agree that the liability of A and B to C shall be joint only, or both joint and several. As joint contractual obligors can be treated as liable severally from the procedural point of view, the addition of several liability in a contractual setting is a clear indicator of an intention to create several liability in the substantive sense.
[91] In the present case Parliament has, in s37(6) of the Securities Act 1978, created a liability which is described as joint and several. The nature of the obligation thereby created is essentially that of a statutory debt. The liability is to “repay”. Conceptually such liability is more akin to a liability in contract than to a liability in tort. By parity of reasoning with the contractual position and s15 of the Partnership Act 1908, I consider the several liability created by Parliament in s37(6) must have been intended to create a liability which was several in the substantive sense. The words “and several” would have been unnecessary if they were intended to apply only in a procedural sense, ie. to signify that the substantive obligation was only joint but was enforceable severally. Indeed that would be so even if the closer analogy were to tort. Furthermore Parliament is unlikely to have used the words “and several” in a limited procedural sense; that would be a misleading way of making what is in any event an unlikely distinction.
[92] Before parting with this case, I wish to mention the decision of the High Court of Australia in Thompson v Australian Capital Television Pty Ltd [1996] HCA 38; (1996) 186 CLR 574. The Court held that the common law rule that a cause of action against joint tortfeasors is one and indivisible had been abolished by a statutory provision corresponding with s17(1) of our Law Reform Act 1936 which provides that judgment recovered against one joint tortfeasor “shall not be a bar” to an action against any other tortfeasor. The provision is therefore couched on the basis of removing a bar, rather than as altering the nature of the cause of action. The High Court of Australia came to the conclusion that the removal of the bar had the effect of turning the previous single and indivisible cause of action into two or more separate causes of action against the individual joint tortfeasors. It is inherent in this view that the section effected a change to the cause of action from its outset, and was designed to do more than simply removing the consequences of the unitary nature of the cause of action as manifested in the judgment rule.
[93] The common law rule was based on the premise that the single and indivisible cause of action which existed in the case of joint tortfeasors merged with the judgment when entered. Although such merger resulted in the cause of action no longer existing as such, the rule in Brinsmead v Harrison (1872) LR 6 CP 584, as it became known, proceeded on the basis that as a result of such merger a judgment against one joint tortfeasor was a bar to an action against all other tortfeasors upon the same cause of action. This is the source of the reference in the statute to removal of the bar. The merger rationale was not, however, the basis for the release rule. That rule applied on the footing that there was no judgment and was based solely upon the unitary nature of the cause of action which, once released as against one tortfeasor, was regarded as having been released as against all.
[94] In the terminology of the times (see Earl Cairns L.C. in Kendall v Hamilton (supra) at 518), entry of judgment was seen as a bar to further proceedings on the basis of the doctrine of merger, but a release was not seen in terms of a bar but rather in terms of the cause of action no longer existing. I do not find persuasive the view that the statutory removal of the bar in the case of judgments has had the effect of also abrogating the release rule. That can hardly have been Parliament’s intent. If it had been, the release rule would undoubtedly have been directly addressed. It is also difficult, when regard is had to the relevant terminology, to see how the removal of the bar can be regarded as having changed the essence or nature of the cause of action. Such removal simply removes one consequence of the unitary nature of the cause of action, namely that a judgment, even if unsatisfied, bars further proceedings upon the cause of action by others jointly entitled to it.
[95] There is in any event a material difference between a judgment and a release. A judgment may be obtained without being of value. A release of its nature will almost always be given for value. If the consideration for it wholly fails, or is not provided, there is no release effective in law and therefore no destruction of the cause of action. There was therefore good reason for the legislature to have distinguished between judgments and releases. It is, for this reason also, inappropriate to conclude that by enacting s17(1) of the Law Reform Act 1936 Parliament unwittingly abolished the release rule as well as the judgment rule.
[96] I am unable to see the decision of the Privy Council in Wah Tat Bank Ltd v Chan [1975] AC 507 as support for the conclusion to which the High Court of Australia came. Contemporary writings did not ascribe this effect to the English equivalent of our s17(1): see McElroy & Gresson: The Law Reform Act 1936 published in 1937 at 81-82 and Glanville Williams: Joint Torts and Contributory Negligence published in 1951 at 44. Nor did Diplock LJ in Bryanston Finance Ltd v de Vries [1975] QB 703, albeit Lord Denning M.R. appears to have done so.
[97] In his judgment in Thompson, with which Gaudron J agreed, Gummow J said at 613:

“At common law, the liability of joint tortfeasors was said to be joint and several. The plaintiff might sue any one of the joint tortfeasors separately for the full amount of the loss, or all of them jointly in the same action.”

If His Honour meant to convey the view that at common law the liability of tortfeasors was substantively both joint and several as opposed to procedurally so, I must respectfully disagree. If such liability was substantively several as well as joint, the cause of action could not have been regarded as single and indivisible, and neither the judgment rule nor the release rule would have had any rationale. I note that Glanville Williams, in the passage referred to earlier, also speaks of joint tortfeasors as being jointly and severally liable without making the distinction between substance and procedure; but the context of his remarks suggests that the several dimension was procedural only. Indeed if that were not so the foundation for the common law rules under discussion would be removed.

[98] The decision in Thompson was reached without reference to the contrary decision of this Court in Brooks v New Zealand Guardian Trust Co Ltd [1994] 2 NZLR 134 (Cooke P, Richardson and Casey JJ). Cooke P, in delivering the judgment of the Court, traversed a number of points which together provide a powerful case that s17(1) was neither intended to have nor has the effect of abrogating the release rule as well as the judgment rule. The point does not have to be formally decided in this case in the light of our conclusions about the proper construction of s37(6) of the Securities Act 1978. Unless and until changed by statute or expressly departed from by this Court in a case where the point directly arises, the decision in Brooks should be regarded as correctly stating the law in New Zealand.
[99] I return to the circumstances of the present case. The settlement entered into by the appellants with Fortex did not address the position of the respondent directors. For the reasons given by Blanchard J such settlement cannot therefore be said to have released them from the several obligations imposed upon them by s37(6). I agree the appeal should be dismissed with costs as indicated.

Solicitors
Chapman Tripp Sheffield Young, Wellington for Appellants
Gilbert Swan, Wellington for Respondents