HIGH COURT OF AUSTRALIA GLEESON CJ, McHUGH, GUMMOW, CALLINAN AND HEYDON JJ CSR LIMITED & ANOR APPELLANTS AND JARRADD EDDY AS ADMINISTRATOR AD LITEM REPRESENTING THE ESTATE OF THE LATE JOHN THOMPSON RESPONDENT CSR Limited v Eddy [2005] HCA 64 21 October 2005 S523/2004 & S524/2004 ORDER 1. Appeal allowed in Matter No S523 of 2004. 2. Appeal dismissed in Matter No S524 of 2004. 3. Set aside order (1) of the Court of Appeal of the Supreme Court of New South Wales dated 26 November 2003 and, in its place, order that: (a) the appeal to that Court is allowed; and (b) the judgment of the Dust Diseases Tribunal of New South Wales dated 4 April 2003 be reduced to $300,419.49. 4. Appellants to pay the costs of the respondent in this Court. On appeal from the Supreme Court of New South Wales Representation: D F Jackson QC with T G R Parker for the appellants in both matters (instructed by Makinson & d'Apice Lawyers) M J Joseph SC with F L Austin and S P W Glascott for the respondent in both matters (instructed by Alex Stuart & Associates) Notice: This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports. CATCHWORDS CSR Limited v Eddy Negligence – Damages – Asbestos-related disease – Compensation for plaintiff's inability to provide domestic assistance to wife – Whether damages are recoverable where a personal injury prevents a plaintiff from providing gratuitous personal or domestic services for another person ("Sullivan v Gordon damages") – Whether Sullivan v Gordon damages are analogous to Griffiths v Kerkemeyer damages – Whether Sullivan v Gordon should be accepted as part of the common law of Australia – Whether Sullivan v Gordon damages could be recovered for those years in which services may have been provided after the plaintiff's death up until the expected date of death but for the tort. Costs – Resolution of legal point – Relevance of recurrent litigant – Relevance of plaintiff with no interest in legal position beyond litigation. 1. GLEESON CJ, GUMMOW AND HEYDON JJ. The defendants (who are the appellants in this Court) admitted liability for negligently exposing the plaintiff to asbestos and thereby causing him to contract mesothelioma. The trial judge, Judge O'Meally, the President of the Dust Diseases Tribunal of New South Wales, ordered the appellants to pay $465,899.49 in damages to the plaintiff[1]. The New South Wales Court of Appeal dismissed an appeal against that judgment with costs[2]. 2. The appellants now appeal against those orders[3]. The point at issue relates to one element in the damages: the figure of $165,480. That sum was described in the courts below as "Sullivan v Gordon damages"[4]. It was awarded as compensation for the plaintiff's inability, after the onset of mesothelioma, to continue to provide domestic assistance to his wife, who suffered from osteoarthritis. She found it difficult to bend and twist and thus to do housework and gardening. Before the onset of mesothelioma, the plaintiff had helped with vacuuming, cleaning, gardening and general maintenance, so far as a pre-existing injury permitted. At the time of the trial in 2003 the plaintiff was aged 61, and it was agreed that he was expected to die in 2004. The plaintiff's wife was aged 60. The figure of $165,480 was calculated on the basis that services would have been rendered for another 20 years, that the plaintiff would have rendered them for one and a half hours a day, and that the cost was $25 per hour. The product of that calculation was discounted by 20 percent for contingencies. No issue is now taken with the correctness of the figure awarded provided the relevant head of damage is recoverable in law, nor, after argument, was any taken at the trial. It is to be noted that the Sullivan v Gordon damages of $165,480 were awarded in addition to general damages of $165,000 and damages under Griffiths v Kerkemeyer[5] of $71,640 for the period before trial and for the period between the trial and the plaintiff's expected date of death[6]. 3. In New South Wales the right to Sullivan v Gordon damages stems from a decision of the Court of Appeal bearing that name[7]. In the present proceedings the appellants challenged the correctness of Sullivan v Gordon at first instance, but the primary judge correctly held himself to be bound by it[8]. The appellants also challenged its correctness in the Court of Appeal, but the Court of Appeal understandably refused to grant leave to re-argue it on the ground that it was a very recent decision by a bench of five judges specially constituted to determine the correctness of a contrary earlier decision of that Court[9]. The issues 1. There are four questions for decision: (a) whether, where a personal injury prevents the plaintiff from providing gratuitous personal or domestic services for another person, the damages recoverable can include an amount calculated by reference to the commercial value of the services; (b) whether that head of damages is recoverable, in the case of an injury leading to death, for the "lost years", that is, in this case, for the 19 years in which the services might have been provided after the plaintiff's actual death up until the date to which he was expected to have lived had the tort not been committed; (c) whether, in the event of the answers to (a) or (b) being favourable to the appellants, the matter should nevertheless be remitted for reconsideration of the figure awarded for general damages; and (d) whether the answer to any of the first three questions should result in alteration to the Court of Appeal's costs order. Each question should be answered in the negative. The state of the law 1. In considering the arguments of the parties, it is convenient to summarise the legal background against which they must be evaluated. 2. First, in Griffiths v Kerkemeyer[10] this Court (Gibbs, Stephen and Mason JJ) held that in a claim for personal injury the plaintiff was entitled to recover an amount equivalent to the commercial cost of nursing and domestic services which had been provided in the past and would be provided in the future by the family or friends of the plaintiff. 3. Secondly, in Van Gervan v Fenton[11] Mason CJ, Brennan, Toohey, Gaudron and McHugh JJ affirmed the view of Stephen and Mason JJ in Griffiths v Kerkemeyer that the true basis of the claim was the need of the plaintiff for the services; that the plaintiff did not have to show that the need was or might be productive of financial loss; and that the plaintiff's damages were not to be determined by reference to the actual cost to the plaintiff of having the services provided or by reference to the income forgone by the provider, but by reference to the cost of providing those services generally in the market. 4. Thirdly, in one jurisdiction there is legislation reversing the rules stated in the Griffiths v Kerkemeyer line of cases[12], and in other jurisdictions there is legislation restricting the availability or the quantum of this head[13]. 5. Fourthly, some jurisdictions, whether by purported application of the rules in Griffiths v Kerkemeyer, or by extension of them, or otherwise, permit recovery of damages reflecting the impaired capacity of plaintiffs to provide domestic services to their families. This claim was rejected in New South Wales by Reynolds and Mahoney JJA (Glass JA dissenting) in Burnicle v Cutelli[14]. That case was followed by a majority (Kennedy and Olney JJ, Wickham J dissenting) of the Full Court of the Supreme Court of Western Australia in Maiward v Doyle[15]. To those jurisdictions can be added Scotland[16]. However, the Queensland Court of Appeal (Macrossan CJ, Davies JA and Fryberg J) accepted the claim in Sturch v Willmott[17]. So did the English Court of Appeal in Daly v General Steam Navigation Co Ltd[18] and the Full Court of the Federal Court of Australia sitting on appeal from the Supreme Court of the Australian Capital Territory[19]. A bench of five members of the New South Wales Court of Appeal in Sullivan v Gordon then adopted a concession by counsel that Sturch v Willmott was correct and Burnicle v Cutelli was incorrect[20]. Since then, Sullivan v Gordon has been followed in Western Australia[21] and the Australian Capital Territory[22]. The opposite view has been taken by a majority of the Full Court of the Supreme Court of South Australia[23]. Principles similar to those stated in Sullivan v Gordon are applied in Canada[24], although there have been dissenting judgments[25]. In the United States there is an avenue of recovery for "a homemaker who is not a wage earner but whose earning capacity is devoted to providing household services"[26]. This was done by extension of principles relating to loss of earning capacity. 6. Finally, the Sullivan v Gordon principle has been assumed, subject to various limitations to be examined later, by the legislatures in Queensland[27] and Victoria[28], and enacted in the Australian Capital Territory[29]. 7. The arguments of the respondent in this Court call for the resolution of the question whether the common law of Australia is reflected in the Sullivan v Gordon authorities. If the answer is yes, the authorities opposed to Sullivan v Gordon must be overruled; if no, the Sullivan v Gordon authorities must be. Sullivan v Gordon as authority 1. The status of Sullivan v Gordon as an authority is heavily qualified by the procedural course which that case took. It was argued on two separate occasions. On the first occasion three judges sat. Neither party referred to Burnicle v Cutelli, although the court evidently did. It directed that the matter be re-listed before five judges for further argument on the correctness of that case. Beazley JA said[30]: "[I]f that decision was to stand, it would govern the outcome of this part of the appellant's claim so as to restrict the plaintiff to having this part of her lost capacity reflected in general damages only. Although making no reference to Burnicle v Cutelli, the trial judge clearly applied the principle stated by the majority that such a claim sounds only in general damages." She then criticised Burnicle v Cutelli, and continued[31]: "On the further argument in the matter, senior counsel for the respondent [defendant] accepted that Burnicle v Cutelli appears no longer to be good law. It will be clear from what I have said that I consider that to be the case." 1. These events placed the Court of Appeal in a difficult position. It is of course commonplace for the courts to apply received principles without argument: the doctrine of stare decisis in one of its essential functions avoids constant re-litigation of legal questions[32]. But where a proposition of law is incorporated into the reasoning of a particular court, that proposition, even if it forms part of the ratio decidendi, is not binding on later courts if the particular court merely assumed its correctness without argument[33]. "[T]he presidents, … sub silentio without argument, are of no moment"[34]. 2. Sullivan v Gordon, of course, cannot be criticised for proceeding by assumption, or sub silentio. The court focused very directly on the question of whether Burnicle v Cutelli should be overruled. But Sullivan v Gordon cannot be regarded as having proceeded by argument. None is recorded, and it flows from the stand of the parties at each of the hearings that very little, if anything, is likely to have been said against overruling Burnicle v Cutelli. No amicus curiae appeared to defend it. The normal function of forensic argument in pointing out difficulties in and necessary qualifications to the competing propositions advanced by adversaries could not be fulfilled. Hence the great advantages of adversarial debate were not available to the court. In Miliangos v George Frank (Textiles) Ltd[35] Lord Simon of Glaisdale said: "although certainly a case is not decided per incuriam merely because it is argued on one side only … , the absence of a contrary argument will sometimes make it easier to establish a per incuriam exception, and in any case a judgment in undefended proceedings or a decision on an uncontested issue tends to have less authority than one given after argument on both sides." Further, Sullivan v Gordon, unlike other cases in which binding authorities have been overruled despite the absence of adversarial argument[36], was not a case in which there was any dissenting judgment which the majority reasoning might have taken account of and profited from. 3. However, whatever the weaknesses of Sullivan v Gordon as an authority, the essential questions must turn on the strength of the reasoning employed in it and cases like it. In this appeal that reasoning was both attacked and defended. Special and general damages 1. There is one aspect of Burnicle v Cutelli which should be noted. The majority did not deny that the lost capacity of injured plaintiffs to assist their families was compensable: it merely said that, if the loss was to be compensated, compensation was to be given not as special damages but as part of general damages. Thus Reynolds JA said[37]: "[A]n assessment must be made as a component of an award of general damages, just as must be done in respect of any other deprivation which does not produce financial loss. The injured plaintiff has in such a case as this lost part of a capacity, the exercise of which can give to her pride and satisfaction and the receipt of gratitude, and the loss of which can lead to frustration and feelings of inadequacy." And Mahoney JA spoke to the same effect[38]. The effect of their reasoning was to deny that compensation for this type of loss was to be calculated by reference to the market cost of replacing the services[39]. Hence when those who support Sullivan v Gordon say that an injured plaintiff who loses the ability to care for a disabled relative loses "something of real value" to the plaintiff as well as the relative[40], they are saying something true, but inconclusive: there is a loss, but it can be compensated as part of general damages. It does not follow from that fact that general damages will compensate for all aspects of the loss of capacity. Nor does it follow that the value of the plaintiff's loss of capacity can be measured by the cost of obtaining care for the disabled relative from professionals. Does Sullivan v Gordon follow from Griffiths v Kerkemeyer? 1. The respondent contended both that Sullivan v Gordon fell within the rules in Griffiths v Kerkemeyer and that it was supportable by analogy with those rules. 2. The appellants' arguments turned on the proposition that either the principle in Sullivan v Gordon fell within Griffiths v Kerkemeyer or it did not; if it did not, the authorities asserting it had effected an anomalous and unwarrantable extension of the law and should be overruled. The proposition was structured in that way because of the majority reasoning in Sullivan v Gordon. 3. The reasoning in Sullivan v Gordon. In Sullivan v Gordon the majority reasoning is that of Beazley JA, with whom Spigelman CJ, Powell and Stein JJA concurred. Perhaps because the party in whose interest it was to defend Burnicle v Cutelli conceded that it was not good law, and perhaps because none of the five judges dissented, the majority reasons were brief. After setting out the authorities, Beazley JA referred to the following statement of Reynolds JA in Burnicle v Cutelli[41]: "There are two losses: one to the recipients of the services and the other to the plaintiff personally. It is easy to quantify the losses to the recipients as being the value to them of the lost services. The difficulty lies in seeing how in principle the loss to [the plaintiff] is to be measured in the same way as the loss to the recipients. I am unable to find assistance in … Griffiths v Kerkemeyer … which … deals with another question, the compensation of an injured person in whom has been created a need, the satisfaction of which calls for the provision of services for which it is reasonable to pay." Beazley JA said[42]: "… I cannot see any logical basis for the distinction drawn in Burnicle v Cutelli between the measure of damages in a traditional Griffiths v Kerkemeyer claim and the measure of damages for the loss sustained by the inability to care for a dependant child. The decision does not, in my opinion, reflect the true nature of a claim of this type, based as it is in a loss of pre-accident capacity which gives rise to a specific post-accident need". She continued[43]: "A person who has lost the capacity to care for a child or children is entitled to be compensated on the same basis as a traditional Griffiths v Kerkemeyer claim." This reasoning plainly views the Sullivan v Gordon head of recovery as based on the rules in Griffiths v Kerkemeyer. Similar reasoning appears to underlie the first reason which the fifth judge, Mason P, gave for overruling Burnicle v Cutelli: he referred to the "plaintiff's accident-created need", being a need to care for other members of the family which was not distinct in principle from the need to care for themselves[44]. 1. The actual basis of Griffiths v Kerkemeyer. However, the Griffiths v Kerkemeyer line of cases does not turn on a "post-accident" or an "accident-created need" in the abstract. In Van Gervan v Fenton[45] Mason CJ, Toohey and McHugh JJ said: "the true basis of a Griffiths v Kerkemeyer claim is the need of the plaintiff for those services provided for him or her". That passage was concurred with by Brennan J[46] and quoted with approval by Gaudron J[47]. When later in their judgment Mason CJ, Toohey and McHugh JJ referred to "need", it was to "need" in that sense. Thus they immediately thereafter asserted the proposition that "it is the need for the services which gives the plaintiff the right to an award for damages"[48]. They reiterated it later when they spoke of "the services required by the injured person"[49] and "the services which the plaintiff reasonably needs"[50]. Although Dawson J did not agree with the majority's approach in Van Gervan v Fenton, he accepted in Kars v Kars that the basis of Griffiths v Kerkemeyer was that a "plaintiff receives the value of services voluntarily provided by way of damages as compensation for the loss suffered by reason of the injuries which manifests itself in the form of a need for those services", and what was in issue was "the voluntary provision of services to a plaintiff"[51]. The majority in Kars v Kars (Toohey, McHugh, Gummow and Kirby JJ) described the principle as permitting recovery of damages "in respect of the cost to a family member of fulfilling the natural obligations to attend to the injuries and disabilities caused to the plaintiff by the tort."[52] The later reference to "the injured plaintiff's … needs" must be understood in the same sense[53]. So must references to the plaintiff's needs in Grincelis v House[54]. 2. In short, as the appellants submitted, Griffiths v Kerkemeyer damages are awarded to plaintiffs to compensate them for the cost (whether actually incurred or not) of services rendered to them because of their incapacity to render them to themselves, not to compensate them for the cost of services which because of their incapacity they cannot render to others. In each instance there may be a "need" for services, but it is a different kind of need, and the recipient of the services is different. 3. The respondent criticised the argument advanced for the appellants by saying that it falsely characterised the plaintiff's reduced capacity to help his wife as causing a loss to his wife, but not to him. That criticism may be fair up to a point, because the reduction in capacity is itself a loss to the plaintiff. But the criticism is not a valid criticism of the proposition that the plaintiff's loss is different from the loss remedied by the rules in Griffiths v Kerkemeyer. 4. It follows that so far as the reasoning in Sullivan v Gordon rested on the view that a Sullivan v Gordon claim has the same basis as a Griffiths v Kerkemeyer claim, it is erroneous[55]. It cannot be said that the Sullivan v Gordon problem falls within the rules stated in the Griffiths v Kerkemeyer line of cases, or within any proposition logically deducible from those rules. 5. Can the outcome in Sullivan v Gordon be supported in any other way? Two ways were suggested – reasoning by analogy with Griffiths v Kerkemeyer, and "policy" reasoning. Is there an analogy between Sullivan v Gordon and Griffiths v Kerkemeyer? 1. The respondent relied on an "analogy with Griffiths v Kerkemeyer". For the following reasons no analogy should be drawn. 2. Controversial character of Griffiths v Kerkemeyer. First, the principle of Griffiths v Kerkemeyer is controversial, as evidenced by the number of legislative reversals or qualifications of it. There is also judicial dissatisfaction with it[56]. It can produce very large awards – some think disproportionately large compared to the sums payable under traditional heads of loss. 3. Anomalous character of Griffiths v Kerkemeyer. Secondly, Griffiths v Kerkemeyer is anomalous in departing from the usual rule that damages other than damages payable for loss not measurable in money are not recoverable for an injury unless the injury produces actual financial loss. 4. A plaintiff who has suffered negligently caused personal injury is traditionally seen as able to recover three types of loss. 5. The first covers non-pecuniary losses such as pain and suffering, disfigurement, loss of limbs or organs, loss of the senses – sight, taste, hearing, smell and touch; and loss of the capacity to engage in hobbies, sport, work, marriage and child-bearing. Damages can be recovered in relation to these losses even if no actual financial loss is caused and even if the damage caused by them cannot be measured in money. 6. The second type of loss is loss of earning capacity both before the trial and after it. Although the damages recoverable in relation to reduced future income are damages for loss of earning capacity, not damages for loss of earnings simpliciter, those damages are awardable only to the extent that the loss has been or may be productive of financial loss[57]. Hence "the valuation of the loss of earning capacity involves the consideration of what moneys could have been produced by the exercise of the [plaintiff's] former earning capacity"[58]. 7. The third type of recoverable loss is actual financial loss, for example, ambulance charges; charges for medical, hospital and professional nursing services; travel and accommodation expenses incurred in obtaining those services; the costs of rehabilitation needs, special clothing and special equipment; the costs of modifying houses; the costs of funds management; and the costs of professionally supplied home maintenance services. It is not necessary for the costs actually to have been incurred by the time of the trial, but it is necessary that they will be incurred. The traditional view of the law was stated by Dixon CJ in Blundell v Musgrave[59]: "[B]efore a plaintiff can recover in an action of negligence for personal injuries an item of damages consisting of expenses which he has not yet paid, it must appear that it is an expenditure which he must meet so that at the time the action is brought, though he has not paid it, he is in truth worse off by that amount. Generally speaking the question whether he must meet the expense is to be decided as a matter depending upon his legal liability to pay it." That principle permits recovery by plaintiffs who, though at the time of the trial they have not made contracts for the provision of future care, will have to do so in future if they are to receive it[60]. Dixon CJ appears to have accepted this in a later statement in the same case[61]: "[T]he basis on which a plaintiff recovers expenses as special damages is that he will have to pay them whether he obtains the amount from the defendant as damages or not." Dixon CJ was in dissent, but not on these principles; Fullagar J (also dissenting) agreed with them[62]. The majority (McTiernan, Williams, Webb and Taylor JJ) assumed them to be sound, differing from the dissenters only on whether there was in that case an obligation to pay. So far as Griffiths v Kerkemeyer permits plaintiffs to recover the costs of nursing and home care services which are to be paid for, it accords with these principles. So far as it permits recovery of those costs even though the services may never be supplied or may never be paid for, it is not only exceptional, but anomalous. 1. Different fields of operation. Thirdly, there is an important difference between the field in which Griffiths v Kerkemeyer applies and the field in which Sullivan v Gordon applies. In applying Griffiths v Kerkemeyer it is relatively easy to estimate the extent of the plaintiff's needs for personal care or services, and to calculate, by reference to the costs of professionals providing that care or those services, what the damages should be (even if it is possible or likely that the care will not be provided, either at all or by paid professionals). But the "need" of the plaintiff to care for others is much harder to evaluate. To examine it by reference to what care the plaintiff ought to have provided in the past would trigger invidious inquiries. To examine it by reference to what care the plaintiff in fact provided in the past would require investigation as to whether the intensity of the plaintiff's interests in providing the services might have been likely to change after the tort because of possible future events like divorce or the birth of new children, or for other reasons. The Sullivan v Gordon problem is not the practical one of calculating costs. It is the legal problem of deciding what test should be employed in deciding what costs need to be calculated. To that Sullivan v Gordon problem there is no Griffiths v Kerkemeyer parallel. 2. Different approach to the "lost years". Fourthly, although Sullivan v Gordon as applied by the Court of Appeal in the present case permits recovery for the "lost years", ie the period within which the injured plaintiffs would have provided services but for the shortening of their lives by reason of the tort, there is no equivalent recovery under the rules in Griffiths v Kerkemeyer. This points against the existence of any analogy. 3. Appeal to incongruity. The supposed analogy between Griffiths v Kerkemeyer and Sullivan v Gordon is sometimes derived from a supposed incongruity – permitting an injured plaintiff to recover Griffiths v Kerkemeyer damages for assistance in housework or gardening which the plaintiff cannot engage in, but not to recover Sullivan v Gordon damages for assistance with the care of a disabled relative which the plaintiff cannot engage in. Thus in Lowe v Guise[63] Rix LJ was critical of the position under which "the loss of an injured person's ability to look after the family garden would be compensated, but the loss of his ability to look after his brother would not be". It is in fact far from clear that there is an incongruity in view of the differences between the field in which Griffiths v Kerkemeyer applies and the field in which Sullivan v Gordon applies. In any event there would be no incongruity if recovery under Griffiths v Kerkemeyer and Sullivan v Gordon were limited to costs that would be incurred. But it was no part of the respondent's submission that the law be reshaped along these lines. 4. Conclusion. In Sturch v Willmott[64] Macrossan CJ perceived an analogy between Griffiths v Kerkemeyer awards and awards now referred to as Sullivan v Gordon awards. He said: "The common element between the two factual categories is the disablement of a plaintiff and the justice of arranging assessments so that wrongdoers do not profit through having their damages reduced by the gratuitous efforts of care providers which are not intended to achieve that result." However, he accepted that the analogy was "very general". Indeed the supposed common element is so general that it does not permit the rules in Griffiths v Kerkemeyer to be used as premises for the conclusion reached in Sullivan v Gordon. Griffiths v Kerkemeyer is well-established, no challenge was made to it in this case, and nothing in this judgment is intended to encourage any future challenge. But to borrow the words of Lord Reid in another context, it is in some ways an "undesirable anomaly", and it should not be applied to "any class of case where its use [is] not covered by authority."[65] Is Sullivan v Gordon supportable on "policy" grounds? 1. The respondent's "policy" arguments outlined. The other avenue of support for Sullivan v Gordon relates to arguments based on "policy"[66]. While much of the advocacy for Sullivan v Gordon consists merely of dogmatic assertions and statements of velleity, particularly by law reform agencies[67], the following arguments of the respondent, which to some extent involved "policy" contentions, rose higher than that. Injured plaintiffs who can no longer provide services to others as fully as they did before the tort have suffered loss of or impairment to their capacity to provide those services to others. That loss of or impairment to capacity is capable of valuation by reference to the market value of the services. Hence it is a compensable form of damage[68]. It is a head of damage for which appropriate compensation is not to be found by relying only on recovery for loss of amenities as part of general damages, for commonly supply of the services does not generate the pleasurable feelings often connected with amenities which have been lost[69]. A particular reason why compensation should be permitted for it is that if the work is not done, the health and safety of families will suffer, and if compensation is refused, the injured plaintiff's family will suffer hardship[70]. The respondent also argued that the authorities opposed to Sullivan v Gordon were flawed by two fallacies. One was that they concentrated too much on the loss to the persons assisted by the injured plaintiff's pre-accident activities, and not enough on the loss of the plaintiff's capacity. The other was that, by stressing the voluntary nature of the plaintiff's pre-accident activities, they obscured the fact that the plaintiff's capacity had economic value. 2. "Policy" arguments in the cases. These and related arguments are commonly employed by academic lawyers and law reform agencies but do not lack judicial support. Thus in Sturch v Willmott[71] Davies JA said: "There are, however, strong policy reasons in favour of measuring the damages in cases … of loss or diminution of capacity by a spouse/parent to provide domestic services formerly provided by her … to her spouse/children, by reference to the commercial replacement cost."[72] He gave four. "One is … that otherwise the wrongdoer may be advantaged at the expense of the plaintiff or her gratuitous helper."[73] He put the second thus: "There are also policy reasons which favour placing an economic value on the domestic contribution of a spouse to her family and treating the loss or diminution of the capacity to make that contribution as the spouse's loss rather than, as in former times, her husband's."[74] The third reason was: "[I]n some cases it may be impossible to disentangle the services which the plaintiff formerly provided to her family from those which she provided for herself"[75]. Fourthly, he said that the loss or diminution in a plaintiff's capacity to provide domestic services was[76]: "analogous to a loss or diminution of earning capacity and should ordinarily be measured by the replacement cost of the services which, by reason of her loss or diminution, the plaintiff is no longer able to provide." 1. It is proposed now to identify some consequences of the contentions thus summarised; to examine the extent to which Sullivan v Gordon has received statutory recognition; to notice some difficult features of the authorities on which the respondent relied; and to explain why, if the principle in Sullivan v Gordon is thought desirable, its introduction should be left to the legislature. 2. The exceptional character of Sullivan v Gordon. It is a general principle of the law relating to the recovery of damages for negligently inflicted personal injury that if the negligence has caused financial loss, it is recoverable as special damages, and if it has caused non-financial loss, that loss is recoverable as a component of an award of general damages[77]. While the courts (whether trial is by jury or by judge alone) take into account the particular ways in which incapacities are said to have revealed themselves after the injury, they do not, unless they are for some good reason specifically requested to do so, fragment the damages in the sense of arriving at separate conclusions about the amount of damages to be awarded for the loss or impairment of each particular capacity. Plaintiffs who, by reason of their injuries, lose the capacity to participate in social or sporting or religious organisations, thereby incidentally ceasing to provide services to those organisations, are compensated for that loss through an undifferentiated element of general damages. So, independently of Sullivan v Gordon, are plaintiffs who lose the capacity to perform domestic services. The effect of Sullivan v Gordon, on the other hand, is that it separates off one aspect of the post-injury diminution in the capacity of plaintiffs and holds that compensation for that particular diminution is not to come as part of a global award of general damages, but by a specific head of special damages. There is no other instance where the diminished capacity of an injured plaintiff is compensated by special damages apart from the exception of Griffiths v Kerkemeyer itself, and the quasi-exception of compensating for diminished capacity to earn by calculating the present value of the future income stream of plaintiffs, usually by reference to their earnings at the time of the accident. It may not matter, except in order to preserve continuity with traditional linguistic usages, whether the issue is posed as being one turning on what damages are recoverable as general damages and what are recoverable as special damages. The substantive issue is, assuming impairment of capacity, how the damages for that impairment are to be assessed. The question is whether there are good legal reasons to select as the basis of calculating damages for the plaintiff's impaired capacity to care for others the sums which those others would have to pay in the market to get the same care. 3. A sub-question, to which there is diversity of response, is how far the impairment of the plaintiff's capacity to provide services is a loss to the plaintiff and how far it is a loss to the recipient of the services. "Doubtless it is more blessed to give than to receive, but surely, when services are terminated, their loss is felt by the person who received them rather than by the person now unable to give them. To say, contrary to the fact, that the loss is the giver's may give rise to a problem of valuation: the value of not having a gift to give may be quite different from its value to the recipient if you gave it."[78] 1. Sullivan v Gordon recovery as a loss of earning capacity. In Sturch v Willmott[79] Davies JA drew an analogy between the plaintiff's loss of capacity to provide services to others and the plaintiff's loss of capacity to earn. If that analysis is sound, it highlights how Sullivan v Gordon recovery operates as an exception to the general rule[80] that damages for loss of earning capacity are only recoverable to the extent that that loss was or might be productive of financial loss. On general principle, if a salaried ambulance worker and a volunteer ambulance worker are injured by the same tort which impairs their capacity to perform ambulance work, the former can recover damages calculated by reference to the probable earnings of ambulance workers, but not the latter[81]. Recovery by an unpaid supplier of domestic services of the commercial cost of replacing those services is a striking exception to general principle. 2. Sullivan v Gordon and losses to the family. The respondent's "policy" arguments for Sullivan v Gordon recovery stressed the importance of that recovery as a means of avoiding "loss to the family"[82]. The law of tort concentrates on compensating injured plaintiffs. Analysing recovery by injured plaintiffs as a means of avoiding loss to their families highlights another exceptional aspect of Sullivan v Gordon. The Scottish Law Commission has put a justification for this approach[83]: "Within the family group, for practical reasons, a system of division of labour and pooling of income obtains in which, though in law the [plaintiff's] services are rendered gratuitously, they are in practice a species of counterpart for the benefits which that member receives as a member of the family group. If by reason of an accident a member of the family group loses the ability to offer the appropriate counterpart for the benefits he receives, he should be compensated for this loss." This is elegant reasoning, but it reveals how remote the loss compensated is from conventional loss and how in substance it is the family which benefits from the award of compensation for the loss even though the "family" is not the plaintiff. 1. Sullivan v Gordon as an indirect benefit to recipients of care. This in turn reveals another principle to which Sullivan v Gordon recovery would be an exception. The understanding underlying the award of Sullivan v Gordon damages is that it gives the plaintiff the opportunity to acquire at commercial rates the services which the plaintiff rendered in the past so that the advantages which the recipients of the services enjoyed in the past will continue in the future. It is true that the plaintiff is not obliged to do that, any more than plaintiffs who recover Griffiths v Kerkemeyer damages for the future are obliged to spend them on securing the provision of care for themselves, for in Australian law there is no trust affecting those damages[84]. But the understanding that injured plaintiffs or persons acting on their behalf will arrange for the services to be acquired will no doubt be fulfilled in many cases. In that practical sense, Sullivan v Gordon awards benefit not only the plaintiff but also the persons cared for. 2. The recovery of the market value of the services which plaintiffs can no longer supply to others creates an indirect avenue of compensation to the persons no longer supplied. The common law gave only limited direct avenues of recovery to those who have lost the benefit of an injured plaintiff's services: the husband's action per quod consortium amisit[85]; the employer's action per quod servitium amisit[86]; and the torts of seduction, enticement and harbouring, by which a father could recover for the loss of his daughter's domestic services[87]. These avenues are now sometimes seen as "antique"[88]. Hence the existence of the husband's action for loss of consortium was not held to justify recognition of an equivalent action in wives[89]. Although the action per quod consortium amisit has been extended by statute in South Australia[90] and has been recognised and modified in Queensland[91], it has been abolished[92] or radically limited[93] in most jurisdictions. The torts of seduction, enticement and harbouring have been abolished in South Australia, the Australian Capital Territory and England[94] and for decades have not been relied on elsewhere. The action per quod servitium amisit has been abolished in England[95] and in large measure in Victoria[96] and the Northern Territory[97]. There are admittedly statutory exceptions to the ban on recovery by those who have lost the services of the deceased, but they are carefully confined[98]. 3. In the United States, the respondent submitted, Sullivan v Gordon damages were recoverable by reason of a widening of the action per quod consortium amisit[99]. It does not follow from the action per quod consortium amisit, by which a husband recovers for loss of his wife's services, that the husband should be able to obtain compensation for his failure to provide services to the wife (or vice versa). And, in view of the varied, but generally hostile, legislative response to the action per quod consortium amisit in Australia, it would not be right to extend it. 4. There are reasons for not creating a further exception to the common law position denying direct recovery to those who have lost the benefit of an injured plaintiff's services by giving them indirect benefits via Sullivan v Gordon. 5. Where a tort causes a supplier of services to cease to supply them to the recipients, to prohibit recovery (subject to very limited exceptions) by the former recipients of the commercial value of the services, while permitting recovery by the supplier of that value, would be anomalous, even if it were intrinsically desirable. The permission would in a practical sense circumvent the prohibition, and would swamp the exceptions to the prohibition. It would cut across the common law refusal to allow a wife to sue for loss of the services her injured husband provided to her, for by allowing the injured husband to recover moneys for the commercial value of those services, it would ensure that normally the wife would enjoy in a practical sense the value of the award. 6. A more specific anomaly would arise on the facts of this case, which took place in South Australia, if the litigation had taken place in a South Australian court. The extension by the South Australian legislature of the action per quod consortium amisit to wives marks the limit which it sees as appropriate for recovery by recipients of services from injured plaintiffs. It is questionable whether the common law in these circumstances should be extended beyond that limit in the manner recognised in Sullivan v Gordon. The extension might also lead to a risk of double recovery: the defendant would have to pay an injured husband the market value of the services which that husband could not provide in future, while also paying the wife what she has lost under the statutory cause of action. There is no simple way of avoiding this outcome unless the unattractive course were taken of reading words into the South Australian statute, or unless a further qualification were imposed on Sullivan v Gordon recovery. The second possibility provides a further argument against recognising Sullivan v Gordon recovery in the first place. It is not surprising that the Supreme Court of South Australia has refused to recognise Sullivan v Gordon recovery[100]. 7. Statutory modification of Sullivan v Gordon. A possible ground for not overruling Sullivan v Gordon might exist if it had achieved certain types of legislative recognition. An example would arise if the legislatures had enacted legislation which assumed its existence and correctness, particularly if the legislation was only workable on the assumption of its correctness[101]. 8. Only three Australian legislatures have dealt with the problem to which Sullivan v Gordon relates. 9. In 2003 the Queensland Parliament enacted s 59(3) of the Civil Liability Act: "Damages are not to be awarded for gratuitous services replacing services provided by an injured person, or that would have been provided by the injured person if the injury had not been suffered, for others outside the injured person's household." Section 59(3) assumes that at common law damages are available for services provided by the injured person – a sound assumption in Queensland if Sturch v Willmott[102] were correct. Section 59(3) limits recovery to non-gratuitous services outside the household. Section 59(3) is an example of Lord Reid's principle: "the mere fact that an enactment shows that Parliament must have thought that the law was one thing does not preclude the courts from deciding that the law was in fact something different"[103]. Section 59(3) is not an example of legislation unworkable unless an assumption as to the common law is correct[104]. To overrule Sturch v Willmott and Sullivan v Gordon would not make s 59(3), which simply limited the common law rule, unworkable. By itself that legislative decision is not a conclusive or even persuasive guide to the content of the common law; it merely reflects a legislative policy choice. 1. Section 28ID of the Wrongs Act 1958 (Vic) provides: "No damages may be awarded to a claimant for any loss of the claimant's capacity to provide gratuitous care for others unless the court is satisfied that – (a) the care – (i) was provided to the claimant's dependants; and (ii) was being provided for at least 6 hours per week; and (iii) had been provided for at least 6 consecutive months before the injury to which the damages relate; or (b) there is a reasonable expectation that, but for the injury to which the damages relate, the gratuitous care would have been provided to the claimant's dependants – (i) for at least 6 hours per week; and (ii) for a period of at least 6 consecutive months." This is like the Queensland legislation in assuming a common law position but cutting it back in particular aspects. 1. The only other Australian legislative initiative was taken by the Australian Capital Territory. Section 100 of the Civil Law (Wrongs) Act 2002 (ACT) provides in part: "(1) A person's liability for an injury suffered by someone else because of a wrong includes liability for damages for any resulting impairment or loss of the injured person's capacity to perform domestic services that the injured person might reasonably have been expected to perform for his or her household if he or she had not been injured. (2) In an action for the recovery of damages mentioned in subsection (1), it does not matter – (a) whether the injured person performed the domestic services for the benefit of other members of the household or solely for his or her own benefit; or (b) that the injured person was not paid to perform the services; or (c) that the injured person has not been, and will not be, obliged to pay someone else to perform the services; or (d) that the services have been, or are likely to be, performed (gratuitously or otherwise) by other people (whether members of the household or not)." Legislation to this effect was introduced in 1991 as s 33 of the Law Reform (Miscellaneous Provisions) Act 1955 (ACT)[105]. It thus postdates the first English case recognising Sullivan v Gordon recovery in 1980[106]. But it predates all Australian cases recognising it except for two decisions of the Full Court of the Federal Court of Australia on appeal from the Supreme Court of the Australian Capital Territory[107]. Apart from them, in 1991 the Australian common law, as reflected in decisions of intermediate appellate State courts, was against Sullivan v Gordon recovery[108]. Section 33 reflects a legislative decision that the law in the Australian Capital Territory should be as stated in the cases decided on appeal from the Supreme Court of the Australian Capital Territory, not as stated in the State courts. It was a response to a 1986 report of the Australian Law Reform Commission recommending the change as desirable in view of a concurrent recommendation to abolish the action per quod consortium amisit[109]. The legislative assumption thus appears to be that Sullivan v Gordon is not part of the common law, for if it were, the legislation would have been unnecessary[110]. Further, the reasoning lacks force as a justification for effecting judicial, as distinct from legislative, change in the common law, because while the action per quod consortium amisit has been abolished in one jurisdiction and limited in other jurisdictions, it survives to some degree[111]. 1. Even if it were possible to infer from the Queensland and Victorian legislation any relevant legislative assumptions, a central difficulty would remain. The controversy concerns the existence of a common law doctrine. In Australia there is a single common law. If every legislature had enacted legislation assuming the correctness of the Sullivan v Gordon doctrine, that might be a pointer towards its existence as a matter of common law. But there is no consistent pattern of State legislation of that kind[112]. 2. Difficulties in Glass JA's reasoning. Although the respondent relied on the reasoning of Glass JA in Burnicle v Cutelli, he, like Reynolds JA and Mahoney JA, refused recovery. He did so because he saw recovery as dependent on whether there was a "reasonable necessity of providing [the] service [formerly supplied by the plaintiff] at a commercial cost"[113]. The claim was based on the work of the injured plaintiff's daughter in cleaning, cooking and laundering for her mother (the plaintiff), her invalid father, her unemployed brother and her school age sister. "The defendant is entitled to say that the domestic burden she bears could be substantially lightened if her mother, brother or father gave her some assistance and that it is not reasonably necessary to procure her services at a cost."[114] Although that entitlement would not avail the present appellants, this qualification was not supported by the respondent. It does, however, highlight uncertainty about how the Sullivan v Gordon principle is to be defined. 3. Difficulties in Daly's case. The result in Daly v General Steam Navigation Co Ltd[115] is defensible to the extent of its being in part an application of Griffiths v Kerkemeyer. The plaintiff recovered for the cost of employing a person to help her with housekeeping tasks carried out for the benefit of herself and her family, although no such person had been or would be employed, the work in fact having been done by her husband and other relatives. The decision, delivered ex tempore, has the curious feature that recovery was said to be allowable for the future on the basis of the cost of employing someone to do the work, whether or not the plaintiff intended to employ anyone, but for the past only on what was spent in employing someone, or on what earnings the plaintiff's husband lost in giving up work to assist her. This is insupportable, because the two approaches cannot stand together. The English Law Commission thought that the common law would develop so as to eliminate the contradiction[116]. But Potter LJ declined to recognise or bring about that development in Lowe v Guise[117]. 4. Difficulties in Lowe v Guise. Lowe v Guise was decided on assumed facts[118]. The claim in that case, an English case, would have been recognised in Scotland as a result of legislation applying only to Scotland – the Administration of Justice Act 1982 (UK), s 9. The Court of Appeal said that the essential question was whether a claim which would succeed in Scotland had to fail in England because Parliament had not seen fit to enact a similar provision for England. The Court of Appeal was acutely anxious to avoid what it saw as incongruity of outcome, and this was important in causing it to reverse either a contrary decision, or strong dicta, of its own given less than three years earlier in Swain v London Ambulance Service NHS Trust[119]. 5. For Scotland, but not for England, Parliament enacted a form of Griffiths v Kerkemeyer recovery in s 8 of the Administration of Justice Act 1982 and a form of Sullivan v Gordon recovery in s 9. Rix LJ in Lowe v Guise[120] said that the reason was that "Parliament believed that in England the common law had developed or was capable of developing along the lines recommended by" the Pearson Report in 1978[121], the English Law Commission in 1973[122] and the Scottish Law Commission in 1978[123]. But what lines, precisely? The Scottish Law Commission[124] recommended, and Parliament enacted, a statutory duty on the plaintiff to account to the relative who has rendered care to the plaintiff for the damages recovered: s 8(2) of the Administration of Justice Act 1982. The Scottish Law Commission differed on this point from the Pearson Report[125] and the English Law Commission[126]. Thereafter the common law of England relating to Griffiths v Kerkemeyer was altered by the House of Lords in Hunt v Severs[127] by holding that there was a trust over moneys recovered by an injured plaintiff in relation to care to be supplied to the plaintiff, the beneficiary being the carer. That is not a proposition known to Australian law[128]. 6. Rix LJ in Lowe v Guise referred to the view of the English Law Commission in 1999[129] that the loss recoverable under Sullivan v Gordon "should be compensated as a pecuniary loss to the claimant where he or she has paid or will pay for the work to be done, as a loss to the third party where that third party has carried out, and will carry out the work for free, and as an element of non‑pecuniary loss where the claimant has struggled on with the work regardless and will continue to do so." He also referred to that Commission's view that it was unnecessary to recommend legislative enactment of this position because "the common law can be expected to reach" it[130]. He did not refer to the Law Commission's recommendations (which correspond with the Administration of Justice Act 1982, s 9) that where damages are awardable for past work, the plaintiff should be under a personal obligation to account to the person who did the work, but that there should be no obligations of that kind to pay third parties for work done in the future. Indeed he implicitly rejected the latter recommendation, because he said of the injured plaintiff in that case, whose inability to care for his disabled brother in the future as he had in the past was expected to be overcome partly by his mother: "To the extent that his mother has by her own additional care mitigated the claimant's loss, it may be that the claimant would hold that recovery in trust for his mother."[131] 7. There is no persuasive value for Australian courts in Lowe v Guise. That is because its conclusions are arrived at by reasoning which seeks at times to relate itself closely to, and at times departs from, legislation enacted in one part but not another part of the United Kingdom and to unharmonious recommendations by Law Commissions and a Royal Commission in the United Kingdom, particularly where local law reform bodies have not made those recommendations and where the content of those recommendations diverges sharply from Australian law. 8. Where is the line to be drawn? The respondent contended that recovery should be limited by reference to services in the nature of domestic work which the injured plaintiff had been providing before the tort, which the plaintiff was morally or legally obliged to provide, which were reasonably necessary, which were generally recognised as having a commercial value, and which went beyond what was generally provided by persons in the relationship between the plaintiff and those receiving the services[132]. But the respondent did not state a legal principle justifying the extending of recovery to or the limiting of recovery at that point. If what is crucial is impairment in the plaintiff's capacity, a question must arise as to why any of these limitations are to be accepted. How far, then, does the Sullivan v Gordon principle go? To loss of capacity to care for close family members (de jure or de facto), or any family members (de jure or de facto), or foster children, or members of the plaintiff's household[133], whether "immediate" or "extended"[134]; and if to any of these classes, only to dependent members of them, or all members of them? If only to close family members, what is "closeness"? If only to dependent members, what is "dependency"[135]? If the test turns on damage to capacity, why should recovery not extend beyond domestic services? Should it apply beyond domestic services to the wide range of educative services healthy parents supply their children of an academic, sporting or cultural kind? "And if the incapacity to give gratuitous services is a loss to the giver, ought one not to pay the pious spinster whose charitable works are inhibited by injury?"[136] Or should it extend to services provided to friends, or to neighbours? Should it extend to plaintiffs who customarily visited or helped many hospital patients, or old people, or destitute people; or provided volunteer emergency services to others even though they were complete strangers to that plaintiff[137]? Does the test turn on a legal duty to provide services, or on a moral duty to do so, or on what services the plaintiff might reasonably have been expected to perform if there had been no injury[138], or on what services were or might have been expected to have been rendered before the injury[139] or on a mere practice of having provided services? If the injury to the plaintiff causes a loss of capacity which has not been utilised in the past to help others, should that loss be compensated under this head? What inquiry should be made into the likelihood that a capacity which has been utilised to help others before the injury would have continued to be utilised after it? Since that likelihood may vary as between fragile and enduring relationships, and since it may have been likely to diminish as the plaintiff became older or more fatigued, is it open to or obligatory for the court to engage in assessment of whether care would continue to be provided, for how long, and at what level? Is this inquiry to be regarded as invidious[140]? Should the same damages be payable to an injured homemaker who did little housework and fed the family on fast food as to an injured homemaker who spent all day working in the home? Or would an inquiry into the plaintiff's levels of skill in and application to the performance of domestic tasks be invidious[141]? Should the injured plaintiffs be under an obligation to account for any recovery to the persons to whom the plaintiffs can no longer perform services? 9. The respondent argued that his success in this case did not depend on these problems being solved, and that their solution could await other cases. This overlooks the need to identify an underlying principle justifying recovery. 10. There is something to be said for and against each of the possibilities just raised, no doubt[142]. The appellants contended that the particular choices to be made were not compelled either by existing legal principle or by any of the forms of legal reasoning employed in developing the common law. They were policy choices to be made, if at all, by legislatures. 11. Factors relevant to developing the common law. In argument various factors pointing for and against the merits of changing the common law as stated in Burnicle v Cutelli were mentioned. 12. The respondent relied on the following factors as reasons not to return to the New South Wales position in Burnicle v Cutelli: that it was unjust and inconvenient; that it was out of line with many other common law jurisdictions; and that there had been no statutory repeal of Sullivan v Gordon, and some legislative recognition of it. On the other hand, there was no contradictor in Sullivan v Gordon; it is difficult clearly to formulate the principle on which it rests; it cannot readily be fitted coherently into the general law of tort; and there has in the past been considerable judicial disagreement about it. If Sullivan v Gordon were thought to represent a desirable principle, it would be better for that principle to be stated clearly in legislation after reviewing the whole of the relevant field, particularly since to some degree it is not a mere matter of lawyers' law, but raises political issues about the legitimate extent of recovery. These stem from the potential scale of recovery. As increasing numbers of people live to great ages, creating a wider need for care, the question of how far defendants who have tortiously injured the carers of those people should be liable becomes both an important question and a question on which the opinions of citizens may differ sharply. The same is true where a young plaintiff has been caring for a young person[143], so that very large awards might be made in circumstances where there is no guarantee that the care would have continued. 13. The respondent's arguments, then, are not necessarily to be rejected for flaws in the policy reasoning on which they rest; they are to be rejected because they rest on policy reasoning which it is more appropriate for legislatures to weigh than for courts. 14. In these circumstances, if it is desired to confer the rights recognised in Sullivan v Gordon on plaintiffs, the correct course to follow is that taken in the Australian Capital Territory and Scotland: to have the problem examined by an agency of law reform, and dealt with by the legislature if the legislature thinks fit[144]. 15. Conclusion. All the Australian cases supporting Sullivan v Gordon as a principle of Australian common law should be overruled[145]. Does Sullivan v Gordon apply to the "lost years"? 1. Since the Australian common law does not recognise Sullivan v Gordon recovery in relation to the period before the plaintiff's death, it does not do so thereafter either. Remission 1. A question arose whether, if the appellants' arguments succeeded, the matter should be remitted in order to examine whether the amount allowed for general damages should be increased. This question should be answered in the negative. 2. The proposition that a plaintiff whose capacity to assist others before the tort can be regarded as an amenity may recover compensation for loss or impairment of that amenity, as part of general damages, has been long recognised[146]. It was accepted in Sullivan v Gordon[147], Lowe v Guise[148] and Carter v Anderson[149]. This head of recovery was available to the plaintiff at the trial in this case. Either it was relied on before the trial judge when the plaintiff claimed $180,000 general damages or it was not. If it was, it was included in the trial judge's award of $165,000 for general damages. If it was not, there is no reason why the respondent should be given the opportunity belatedly to seek more in a second trial on that question. 3. Further, the amount recoverable, whether or not it has already been recovered, is likely to be relatively small, if only because of the much graver seriousness of the other factors going to the general damages awarded and because of the short period – less than two years – in which the relevant capacity was impaired. There is no injustice in depriving the respondent of the chance to obtain that amount in view of the fact that most of what the plaintiff claimed under Sullivan v Gordon was allowed under Griffiths v Kerkemeyer. 4. The figure of $165,480 was justified in this Court by reference to (inter alia) gardening services, looking after a car, painting the house and maintaining the house. The Court of Appeal saw things differently: for it the figure of $165,480 represented the plaintiff's "loss of capacity to care for his disabled wife"[150]. These words apparently refer to the plaintiff's loss of capacity to assist his disabled wife in cleaning the house. Apart from that, what he lost was the capacity to carry out gardening, maintenance, and mechanical work on the family car. As pleaded, the plaintiff's claim for Griffiths v Kerkemeyer damages and his claim for Sullivan v Gordon damages were partly separate and partly intermingled. Relevantly the statement of claim gave the following particulars: "As a consequence of the injuries above, the Plaintiff has been unable and/or restricted to carry out work in and about his own home which he was formerly able to perform. Further, the Plaintiff has required, and will continue to require care and assistance from family members. Further, the Plaintiff makes a claim for the loss of his capacity to provide services to his household and wife and/or pursuant to the principles set down in Sullivan v Gordon … … These services included vacuuming, mopping the floor and cleaning the bathroom. Prior to his contraction of mesothelioma, the Plaintiff spent, on average, 3‑4 hours each day providing services to his wife and household. Such services included mowing the lawn; edging; maintaining the garden; pruning; trimming trees; hosing the gardens; outside maintenance; cleaning the car; vacuuming; turning mattresses; assisting his wife shop; running errands for his wife; and paying bills." Virtually all of the services were not services to the wife alone, but would also have brought benefits to the plaintiff himself and the other member of the household, the wife's mother. 1. The plaintiff's written submissions to the primary judge contended that under Griffiths v Kerkemeyer the plaintiff was entitled to recover for paid services in relation to gardening, looking after the car and painting the house; while under Sullivan v Gordon the plaintiff was entitled to recover for paid services in relation to cleaning and shopping. Past Griffiths v Kerkemeyer damages were agreed at $21,828 (with interest of $945). Griffiths v Kerkemeyer damages for the future were agreed at $49,812. 2. Assuming that the respondent's characterisation in this Court of the plaintiff's losses is correct, although they cannot be allowed under the Sullivan v Gordon principle because it does not exist, most of them were in fact allowed at trial under Griffiths v Kerkemeyer. Even if that characterisation is not correct, the minds of the parties and the court were sufficiently concentrated by the evidence not merely on the financial consequences of the plaintiff's loss of capacity, but on the fact of that loss of capacity considered as a loss of amenity. 3. In these circumstances it would be a waste of the parties' financial resources and the court's time to engage in a further trial to decide whether general damages should be increased. 4. After the close of oral argument the appellants sought leave to rely on further written submissions which were directed to withdrawing what was described as a concession made in the course of oral argument. That concession went only to the question of whether there should be a remission of the proceedings for further assessment of general damages. Accordingly it is not necessary to consider whether leave to rely on the further written submissions should be granted. Costs 1. The appellants submitted that if the appeal succeeded they should have their costs both in the Court of Appeal and in this Court. 2. The basis of this submission was that the Court of Appeal ordered, in relation to the proceedings before it concerning Sullivan v Gordon, that the appellants pay the costs, and that those costs be assessed on an indemnity basis from 3 June 2003. On that day the plaintiff offered to accept $115,470 in settlement of the appeal. On 17 June 2003, the appellants made a counteroffer of $35,000 which was rejected. The plaintiff's successful resistance to the challenge to Sullivan v Gordon not only caused the appeal to be dismissed, but also meant that the plaintiff had bettered his offer while the appellants had not bettered theirs. The appellants contended that if Sullivan v Gordon were overruled in this Court, the foundation for the costs orders would have disappeared. 3. It is notorious that over many years the first appellant and other members of the group of companies to which it belongs mined asbestos, and manufactured and supplied asbestos-based products. Very large numbers of their employees have been exposed to asbestos; many of them have contracted asbestosis and mesothelioma as a result; many admissions or findings that these diseases were caused by their negligence in this respect have been made; many will be made in future. The appellants challenged Sullivan v Gordon below, applied for special leave to appeal, and prosecuted the appeals, in order to vindicate their long-term commercial interests, for success will unquestionably tend to reduce the quantum of damages payable by them in asbestos-related litigation, of which, unfortunately, there appears likely to be a large quantity in future years. 4. In contrast, the plaintiff had no interest in the legal position beyond this particular litigation. Since the plaintiff's death the same is now true of the respondent, the administrator ad litem. It was entirely reasonable for the plaintiff to seek an award of Sullivan v Gordon damages in the Dust Diseases Tribunal of New South Wales, since that court was bound by the decision of a five-judge Court of Appeal in that case. The challenge to Sullivan v Gordon before the primary judge (which was inevitably rejected), in the Court of Appeal (which was not surprisingly repelled) and in this Court (which has succeeded) made this case a test case, designed to resolve a conflict amongst the intermediate appellate courts of the States and the Australian Capital Territory. It is common in this Court in cases where the resolution of a point is desirable from the point of view of a large and recurrent litigant, whether corporate (for example, an insurance company) or governmental (for example, the Commissioner of Taxation or the Australian Competition and Consumer Commission), but the other party to the litigation is not a recurrent litigant and is not well-positioned to meet adverse costs orders on the point being tested, for the grant of special leave to be made conditional on appellants paying the other side's costs in any event and on appellants not seeking to disturb costs orders in the courts below which were favourable to the other side. At the hearing of the special leave application, the respondent contended that special leave should only be granted on terms of that kind. The application was reserved to be dealt with on the hearing of the appeals. The appellants contend that these terms should not apply because of the costs offers made in June 2003, and because the appeals to this Court were only rendered necessary because of the plaintiff's decision to institute the proceedings in New South Wales (where Sullivan v Gordon applied) rather than South Australia (where it did not). These matters do not make the imposition of the terms requested by the respondent unjust. In the circumstances described above, it is appropriate that those terms as to costs apply. They are reflected in the orders proposed below, paragraph 2 of which will leave the costs orders of the Court of Appeal undisturbed. Orders 1. The following orders should be made: 1. The appeal in relation to the award of Sullivan v Gordon damages (Matter No S523 of 2004) is allowed. 2. The appeal in relation to costs (Matter No S524 of 2004) is dismissed. 3. Order (1) of the Court of Appeal of New South Wales dated 26 November 2003 is set aside and, in lieu thereof, it is ordered that: (a) the appeal to that Court is allowed; and (b) the judgment of the Dust Diseases Tribunal of New South Wales dated 4 April 2003 be reduced by $165,480.00 from $465,899.49 to $300,419.49. 4. The appellants pay the costs of the respondent in this Court. 1. McHUGH J. The principal issue in this appeal is whether a plaintiff may recover damages for the loss of capacity to perform gratuitous work for the benefit of his or her disabled spouse. In Griffiths v Kerkemeyer[151], this Court held that an injured person may recover from a tortfeasor the commercial cost[152] of domestic services provided to him or her to satisfy a need created by the injury, regardless of whether or not those services are provided gratuitously. Central to the issue in this appeal is whether the principle upon which Griffiths v Kerkemeyer was decided applies, or should be extended, to a claim for the costs incurred in replacing the provision of gratuitous services by the plaintiff to a spouse or other person. 2. If that case does cover this class of claim, either directly or by a principled extension, a further question arises: does it apply to a period after the plaintiff's death, so that the plaintiff may recover the cost of providing care to dependants during the "lost years" of his life? The duration of the "lost years" is the difference between the plaintiff's life expectancy prior to the injury, and the actual date of the plaintiff's death. 3. In Sullivan v Gordon[153], the Court of Appeal of the Supreme Court of New South Wales held that the Kerkemeyer principle applied to such cases. In effect, it followed the decision of the Queensland Court of Appeal in Sturch v Willmott[154] where Macrossan CJ recognised that to allow the plaintiff to recover the cost of providing services to others was probably an "extension" of Griffiths v Kerkemeyer[155]. In turn, Sullivan has influenced the development of the law of damages in several States[156]. The New South Wales Court of Appeal applied it in this case after refusing the appellants leave to argue that the Dust Diseases Tribunal of that State had erred in awarding damages in accordance with the decision in Sullivan v Gordon. Statement of the case 1. In September 2002, John Leonard Thompson sued CSR Limited, the first appellant, and Midalco Pty Limited, the second appellant, for damages in the Dust Diseases Tribunal of New South Wales. In the action, he claimed that, as a result of exposure to asbestos dust and fibre, he had developed malignant mesothelioma. The appellants admitted liability. The parties conducted the litigation on the basis that Mr Thompson was expected to die from mesothelioma in February 2004. In the event, he died some months earlier. 2. O'Meally P, who tried the action, awarded Mr Thompson damages of $465,899.49. Included in the award was the amount of $165,480 for Mr Thompson's loss of capacity to care for his disabled wife including care that would be needed after his death. This head of damages was described as the Sullivan v Gordon[157] head of damages. The wife suffered from osteoarthritis and was unable to undertake heavy domestic duties in and around the house. Mr Thompson performed these tasks until he became too debilitated to do so. 3. The appellants appealed to the Court of Appeal of the Supreme Court of New South Wales on the ground that O'Meally P erred in allowing Sullivan v Gordon damages. The Court of Appeal rejected the appellants' application for leave to re-argue Sullivan v Gordon and upheld the award made by O'Meally P. The categories of special damages should not be extended 1. The distinction between general and special damages is central to this case. Until the decision of this Court in Griffiths v Kerkemeyer[158], the Court had employed the traditional usage of the terms. In Paff v Speed[159], Fullagar J defined the difference between them in the following way: "Special damages are awarded in such cases in respect of monetary loss actually suffered and expenditure actually incurred. Their two characteristics are (1) that they are assessed only up to the date of verdict, and (2) that they are capable of precise arithmetical calculation or at least of being estimated with a close approximation to accuracy. The familiar examples are medical and surgical fees paid or payable, ambulance and hospital expenses, and loss of income. Where the plaintiff has been employed at a fixed wage or salary, his loss of income can commonly be calculated with exactness. Where the plaintiff has not been employed, but is, for example, a professional man, his monetary loss can be estimated without difficulty by reference to his past earnings. In a high proportion of cases the amount of the 'special damages' is agreed between counsel for the plaintiff and counsel for the defendant. 'General damages' on the other hand, are, of their very nature, incapable of mathematical calculation, and (although the expression is apt to be misleading) commonly very much 'at large'. They are at large in the sense that a jury has, in serious cases, a wide discretion in assessing them. Also general damages may be assessed not with reference to any limited period, but with reference to an indefinite future. Damages may be awarded for 'pain and suffering', and such damages are assessable for past, present and future pain and suffering. But here calculation is obviously impossible, and damages for pain and suffering should clearly be regarded as 'general' and not 'special' damages. In fact, the question of general damages is generally, I think, put to a jury under three heads – (1) 'economic loss', (2) loss of 'amenities' or 'enjoyment of life', and (3) pain and suffering." 1. On this theory of the difference between general and special damages, then, the claim by Mr Thompson for damages concerning assistance to his wife was one of general damages. In the strict sense, the damages sought were compensation for an immeasurable lost capacity to care for his wife. However, since the decision in Griffiths v Kerkemeyer, the distinction between special and general damages in some cases has been blurred, if not rendered entirely redundant[160]. In contrast to the traditional claim for special damages, the damages awarded in Griffiths v Kerkemeyer were not confined to the date of the verdict. They covered the future. Griffiths v Kerkemeyer also recognised that the relevant loss was the loss to the plaintiff of his capacity, as represented by his need for assistance. The value of that loss was perceived to be the value of the services needed to provide that assistance. Mason J said that[161] "[i]n general the value or cost of providing voluntary services will be the standard or market cost of the services". Hence, it was irrelevant that the need was met by the gratuitous services of a wife or other household member, or indeed, any other person. 2. Griffiths v Kerkemeyer illustrates the truth of Holmes' dictum[162] that the "life of the law has not been logic: it has been experience." As a matter of principle, Griffiths v Kerkemeyer damages are an anomaly. There is no reason in principle why the inability of an injured person to meet his or her needs should be regarded as a special case, no reason why that inability should be distinguished from incapacities such as restriction of use or movement or the pursuit of social, sporting or business activities. Incapacities falling into the latter categories are compensated under the head of general damages. They are compensated in the same way as pain and suffering under the general head of the loss of enjoyment of life. They are not given a special award of damages. In principle, neither should incapacity resulting in the need for services, except in respect of liabilities incurred up to the date of verdict. 3. It is true that, for a long period before Griffiths v Kerkemeyer was decided, common law judges, sitting without a jury, almost invariably calculated any future loss of income as a special head of damages. Even in jury trials – at least in New South Wales – counsel addressed the jury on the basis that the lump sum verdict would include a specific sum for any future loss of wages. This practice was deplored by Barwick CJ in Arthur Robinson (Grafton) Pty Ltd v Carter[163]. His Honour pointed out – correctly in my opinion – that the subject of compensation was "the loss of earning capacity" and that reference to a sum representing the present value of predicted weekly loss of wages was "the merest guide and insusceptible of forming the basis of a calculated amount."[164] 4. Nearly 30 years ago, the thesis of Professor Patrick Atiyah's Inaugural Lecture at Oxford[165] was that, since about 1850, the strict application of general rules to determine legal issues has been in decline. Individualised justice has come to prevail over the application of general rules. If a rule leads to a result in a particular case that is inconvenient or contrary to contemporary notions of fairness or justice, the tendency of the judicial process has been to ignore or distinguish the rule. It is not enough that the rule is sound and applies fairly in the majority of cases. 5. In his Lecture, Professor Atiyah used the term "principle" instead of "rule". But the context makes it clear, as Professor Julius Stone has pointed out, that, when Professor Atiyah used the term "principles", he was using it in the sense of "what we usually call 'rules', that is, precepts prescribing detailed legal consequences for precisely predicated sets of facts"[166]. Professor Atiyah asserted[167] "that the courts have become highly pragmatic and a great deal less principled." His thesis was powerfully criticised by Professor Stone[168]. 6. Whether or not one sides with Professor Stone in this jurisprudential debate – as I do – no one familiar with the course of authority in recent decades could fail to see that courts are far more pragmatic than they once were. If the courts perceive a rule as requiring an unfair or unjust result in a particular case, they are likely to distinguish the rule, make an exception to it or even in some cases abolish it. Courts are much more ready to do this than they once were. Pragmatism has become a powerful force in the law as well as in politics and philosophy. Judge Posner has even argued that all judicial decisions should be based on pragmatism in the sense of practical reasoning[169]. 7. One can accept – as I do – that the first statement of a legal rule is not its final statement and that the utility of the common law requires it to be constantly updated to serve the current needs of society. Nonetheless, the common law would become unpredictable if judges were free to decide cases in accordance with their own notion of what justice requires or what is the most practical way to settle a dispute that comes before the court. So far as is possible, the body of legal rules should constitute a coherent whole. As Lord Devlin pointed out in Hedley Byrne & Co Ltd v Heller & Partners Ltd[170]: "The common law is tolerant of much illogicality, especially on the surface; but no system of law can be workable if it has not got logic at the root of it." 1. One may think that the decision in Griffiths v Kerkemeyer was a just decision that rightly helped to shift the burden from the carers of injured persons to the pockets of the insurers who stand behind most defendants in personal injury cases. Yet even today after many years of applying it, it is not easy to accept that it logically fits in with the principles for assessing damages in such cases. But whether or not that is so, the rule in Sullivan v Gordon is even further removed from a logical application of the principles for assessing damages in personal injury cases. And it is not a logical extension of or valid analogy with the rule in Griffiths v Kerkemeyer. 2. Griffiths v Kerkemeyer was the culmination of a course of authority which had commenced in England and held that an "injured plaintiff can recover the value of nursing and other services gratuitously rendered to him by a stranger to the proceedings"[171]. Initially, it may have been based on the view that, as a matter of ordinary fairness, the person providing the gratuitous services should be reimbursed for the services provided[172]. Lord Denning went so far as to hold that the plaintiff held this part of the damages in trust for the person providing the services[173]. Later decisions seem to have been based simply on the view expounded by Scarman LJ[174]: "The defendants' wrong has created a need for the services. Nursing and attendance are services which can only be provided by expenditure of effort or money, or both: an estimate must be made of the capital value of such effort and money, past and future, and compensation awarded accordingly. How or on what terms they are provided is not of critical importance ..." 1. Similar considerations obviously influenced Stephen J in Griffiths v Kerkemeyer[175] where his Honour said: "The perils of the road, combined with advances in medical knowledge and treatment, ensure that every year a number of road victims survive as helpless invalids, requiring constant attention during many years in the future. For some of them satisfactory care in home surroundings is both possible and adequate, indeed it may have distinct psychological advantages as compared with a life-time in hospital. But it necessarily entails devoted care on someone else's part, often a wife or woman relative who may have to abandon her ordinary employment to nurse the plaintiff and who will in any event find the task a demanding one. In the past it has been customary to disregard the value of such voluntary services when assessing damages in such cases. The result has been to benefit defendants, their insurers and, ultimately, the community at large at the expense of those who, behaving 'like an ordinary decent human being' ... have voluntarily undertaken the care of a loved one maimed on the roads ..." 1. Numerous Australian courts have recognised that the principle established by Griffiths v Kerkemeyer was exceptional in permitting the plaintiff to claim special damages for the loss of capacity to care for him or herself and the resultant need for services from another[176]. But, as members of this Court have pointed out, it is now too late to reverse it by judicial decision[177]. 2. In holding that a plaintiff could recover the value of gratuitous services, Griffiths v Kerkemeyer was bound to unsettle the long established rule[178] that an item of special damages could only be recovered as compensation in respect of a liability paid or incurred. Stephen J recognised this in Griffiths v Kerkemeyer when he noted[179] that "in this particular area of the law [the principle] deprives of all substantive significance the distinction between special and general damages". As a result, Australian courts have extended the Griffiths v Kerkemeyer principle to other cases of gratuitous services which previously would not have been the subject of compensation. They have extended it: . to care provided by a plaintiff-mother to her children[180]; . to cleaning work performed by a plaintiff-wife in her husband's hairdressing salon[181]; and . to the cost of care for a plaintiff-mother's children for the period of time after her injury-caused death until they no longer required such care[182]. 1. The decisions in these cases show that Griffiths v Kerkemeyer has not only unsettled the distinction between general and special damages but it has had unsettling consequences for the legal principles governing the doctrines of loss of consortium and per quod servitium amisit. In Wright v Cedzich[183], this Court and, in Best v Samuel Fox & Co Ltd[184], the House of Lords held that the action for loss of consortium – an action available to a husband – was not available to a wife. In Best, Lord Porter said[185]: "Even if it be conceded that the rights of husband and wife ought to be equalized I agree with the Lord Chief Justice that today a husband's right of action for loss of his wife's consortium is an anomaly and see no good reason for extending it. If the change is to be made I should prefer to abolish the husband's right rather than to grant the like remedy to the wife." 1. In permitting a wife to recover damages for her lost capacity to provide services to her husband, the Sullivan v Gordon extension duplicates the husband's action for loss of consortium and indirectly allows the wife to recover compensation for services which Wright and Best held were not recoverable at common law. Moreover, in allowing the husband to recover, in substance Sullivan revives a cause of action that a number of Australian legislatures have either abolished[186] or limited[187]. Moreover, in permitting a wife to recover for loss of services that she performed gratuitously in her husband's business, the extension of Griffiths v Kerkemeyer has indirectly outflanked the action per quod servitium amisit. That action was confined to the loss of services rendered by an employee[188]. 2. In Sturch v Willmott, Macrossan CJ said that once it was decided that damages were available for services previously provided by the plaintiff to another, it was difficult to "[limit] the ambit of a defendant's liability to compensate."[189] His Honour said[190]: "Many plaintiffs, before being injured, may have been performing voluntary services for persons outside their immediate households and doing so, not in response to anything which could be regarded as necessity, but simply as a result of a free and unpressured choice." 1. The respondent attempted to justify retaining the decision in Sullivan v Gordon by contending that such damages could be limited to the loss suffered by a household. In Sullivan v Gordon[191], Mason P supported this approach saying: "Different considerations probably apply in the case of persons for whom no legal obligation of care exists and who are not members of the plaintiff's household being cared for at the time of the accident (for example, aging parents). The right recognised here almost certainly does not involve exclusively moral obligations of care of persons outside the immediate household ..." 1. But there are many difficulties with the limitation suggested by the respondent. It is an arbitrary limitation that bears no relationship to the principles upon which damages for personal injury are assessed. Given those principles, the suggested limitation lacks a foundation in logic. Why should it be arbitrarily confined to the care of members of the household and not to other close relatives who need care? Why should it be arbitrarily confined to care and not to the need for assistance? And if the need for assistance is the criterion why should the defendant have to pay damages because a wife can no longer work in her husband's hairdressing shop, a class of case that has been held to fall within the Sullivan v Gordon extension? 2. The basis for the limitation appears to be the unstated premise that a loss of capacity for one member of a household is a loss to the entire household because a family is an economic unit. However, this immediately gives rise to the question of how "household" should be defined. Is it limited to legal marriages, or de facto marriages? Does it extend to any person residing in the same dwelling? What of parents who must care for their disabled adult children or grandparents who have spent time in the homes of their children caring for grandchildren who do not live with the grandparents? Why should this head of damages not extend to nieces and nephews taking care of aunts and uncles? Other persons that would seem to have moral claims as strong as those of members of the household include siblings, siblings-in-law, ex‑partners, future children[192] and friends and neighbours. 3. Intuitively attractive though the notion of the household limitation may be as a brake on this head of damage, on examination it is an arbitrary limitation lacking an acceptable basis in legal logic, moral obligation and social policy. The "household" limitation is not a solution to the problem of limiting Sullivan v Gordon. Need and loss of capacity 1. There has been some discussion throughout the cases in this area as to whether the correct conceptual approach is to be found by viewing the compensable loss as a need or as a loss of capacity. In Burnicle v Cutelli[193], Glass JA said[194]: "I am unable to see any reason in point of doctrine why the conceptual approach … adopted in Donnelly v Joyce and Griffiths v Kerkemeyer ... should include a need for nursing services due to an impaired capacity to do for oneself but should exclude the need for domestic services due to an impaired capacity to do for one's family. … Granted that the impairment of a capacity to attend to one's own toilet and similar needs is compensable I am unable to distinguish in point of principle the impairment of the capacity to keep house for one's family. Damages would be recoverable in each case by setting an objective value upon the depreciation of an economic asset. The financial saving involved in the exercise of each of these two capacities is demonstrated by the financial cost which may be entailed by the inability to exercise it." 1. Glass JA dissented in Burnicle v Cutelli, but his views prevailed in Sullivan v Gordon upon a concession by counsel that Burnicle had been wrongly decided. This statement of Glass JA can be accepted so long as the loss of capacity to provide domestic services for one's family is treated as part of the plaintiff's claim for general damages. But it cannot be accepted in so far as Glass JA intended his statement – as he surely did – to justify a separate head of damages equivalent to a claim for special damages. The special head of damages that Griffiths v Kerkemeyer sanctioned was based on need, not loss of capacity. It was the plaintiff's need that was regarded as decisive in Griffiths v Kerkemeyer. In Nguyen v Nguyen[195], Dawson and Toohey JJ and I interpreted Griffiths as holding that "the plaintiff's loss … was represented by [his] need." In Van Gervan v Fenton[196] Mason CJ, Toohey J and I said that passage correctly interpreted the majority view in Griffiths. We also said that that view was consistent with the salient passage of Donnelly v Joyce upon which the majority in Griffiths based their judgment. That passage reads[197]: "The plaintiff's loss, to take this present case, is not the expenditure of money to buy the special boots or to pay for the nursing attention. His loss is the existence of the need for those special boots or for those nursing services, the value of which for purposes of damages – for the purpose of the ascertainment of the amount of his loss – is the proper and reasonable cost of supplying those needs." (emphasis added) 1. Because the costs of care associated with servicing an injury‑caused need are relatively easy to quantify, Griffiths v Kerkemeyer damages have been dealt with as a head of special damages. Yet even if one were to accept that it is the loss of capacity, rather than the specific need, that is relevant, that would not justify establishing a separate head of special damages for the purposes of compensating a Sullivan v Gordon-type loss. An analogy is often drawn between this head of damages and the damages awarded for loss of earnings. The conceptual basis for loss of earnings is undoubtedly the lost capacity to earn. Yet, traditionally loss of earnings is one of the headings of general damages. I have already pointed out that, in Arthur Robinson (Grafton) Pty Ltd v Carter[198], Barwick CJ lamented the fact that there is any tendency to create a head of special damages in relation to lost earnings. 2. However, quantifying loss of earning capacity by reference to the plaintiff's potential loss of earnings is not an analogy that can be validly compared with a claim for Sullivan v Gordon-type damages without a revolution in the principles concerning the assessment of damages for personal injury. The present value of the potential loss of earnings may be regarded as a powerful indicator of the value of the plaintiff's earning capacity in the same way that the present value of the predicted earnings of a business are regarded as fairly reflecting its capital value. As long as the departure from principle involved in assessing claims for loss of earning capacity is treated as a narrow exception, the principles involved in assessing damages for personal injury can be maintained. That could not be the result if every incapacity of an injured person was to be treated as a special head of damages whose value was measured by the market value of the activities that could have been used or supported by that capacity. If that approach were adopted, the concept of a lump sum payment for the plaintiff's injury would be replaced by the market value totals of the actual and supposed capacities of each plaintiff. The incapacity to play golf or go to concerts would be evaluated in terms of what golfers or concert-goers were prepared to pay for the pleasure of their pastimes. Whether or not a plaintiff's lost capacity to care for others is realistically capable of being given a market value, to introduce it as a new category of special damages would distort the long established principles for assessing damages in personal injury actions. 3. If the law of damages is to retain its coherence, overruling Sullivan v Gordon is a necessity. It is a decision inconsistent with long established principle and, with great respect to those who have held to the contrary, it finds no support in Griffiths v Kerkemeyer. The critical difference between Griffiths v Kerkemeyer and the cases that have extended it is that Griffiths v Kerkemeyer damages arise as a direct result of the creation of a need in the plaintiff of the provision of the particular services. It is therefore inherently limited. By contrast, no inherent limit exists for Sullivan v Gordon-type damages. The oft‑used example of voluntary work for Meals on Wheels is but one instance of gratuitous work that is not performed out of any legal obligation, but which nonetheless has a significant economic value. Commentators have pointed out that domestic work is not entirely analogous to this type of civil society engagement because domestic work is not optional[199]. Somebody must do it[200]. This is undeniable. Using the term "voluntary", therefore, is apt to mislead in this context because it focuses attention on the type of work being done and whether it is remunerated. But the correct question is whether the work is being performed in response to an injury-caused need of the plaintiff[201]. This is the essence of the error in Sullivan v Gordon: it moved from the needs of the plaintiff which Griffiths sanctioned to the needs of third parties. It elided the plaintiff's injury-caused needs with the pre‑injury needs of others, albeit using one of the policy considerations behind Griffiths v Kerkemeyer, namely that innocent parties should not suffer unrecoverable losses as a result of the tortfeasor's negligence[202]. However, it is not enough that the policies supporting Griffiths v Kerkemeyer damages and Sullivan v Gordon damages are consistent. Sullivan v Gordon is not supported by anything decided or said in Griffiths v Kerkemeyer. Sullivan v Gordon damages are not concerned with the injured person's needs but the needs of a third party for whom the injured person has provided services. Sullivan v Gordon and the cases that follow it or were decided on the same or similar grounds are wrong in law and must be overruled. 4. Instead of treating Mr Thompson's lost capacity to care for his wife as a special head of damage, the proper approach was to compensate his loss by making provision for it in the award of general damages. His lost capacity was compensable under the headings of loss of amenity, loss of enjoyment of life, and the distress and perhaps the psychological pain and trauma, that he would undergo as a result of his inability to care for his disabled wife. The way that Mr Thompson particularised his losses in his submissions makes this clear. He claimed that his "loss of capacity" included: "(a) an incapacity to garden the family home and to perform general house maintenance. This included, but was not limited to gardening and lawn care, maintenance and operation of the irrigation system, the cleaning of gutters, and the repair and fixing of fences; and (b) an incapacity to undertake mechanical work on the family car, such as the performance of light mechanical repairs, the rotation of the tyres, the changing of oil and oil filters; and (c) an incapacity to assist his wife in undertaking the heavy domestic duties required in and around the house." 1. To the extent that Mr Thompson took pleasure in gardening and attending to the car, he would be entitled to damages for loss of amenity and enjoyment of life. To the extent that his injury prevented him from performing these tasks and necessitated the provision of services from another person, there is no reason why he would not be eligible for Griffiths v Kerkemeyer damages at the market rate for those services. The same is true in relation to the domestic duties that he had performed around the house. As the Federal Court of Australia said in Hodges v Frost[203], "[h]ere, the needs were the commingled needs of husband and wife, but no less the needs of the [husband] because they were in some cases mutual." It is an unfortunate aspect of this case that the case was pleaded on the basis that the domestic work was performed for the benefit of the wife. This is no criticism of the deceased's legal advisers. They relied on Sullivan v Gordon. No doubt the case was pleaded in the way it was because of Mr Thompson's desire to make provision for the care that his wife would require during the "lost years" after his death. But unfortunately the way that the case was pleaded now means that the agreed amount of general damages is lower than the damages he would otherwise have obtained. 2. Given the way that the case was conducted before the Dust Diseases Tribunal and the agreement of the parties as to general damages, it would be contrary to the long standing practice of this and other appellate courts to remit the case for a further trial to decide whether the general damages should be increased. Costs 1. The appellants contended that, if the substantive appeal were allowed, they should have their costs in the Court of Appeal and in this Court. I can see no basis for refusing the appellants their costs of the proceedings in this Court. It is true that the first appellant and its subsidiaries have been sued in a number – probably a large number – of actions for negligence arising out of the manufacture and supply of asbestos-based products. It seems likely – at all events it is quite possible – that success in the substantive appeal will have financial advantages for the appellants that extend beyond those obtained in setting aside Mr Thompson's claim for Sullivan v Gordon damages. Probably, they are defendants in a number of cases in which this type of damages are being, and in the future will be, sought against the first appellant or its subsidiaries. But the appellants' relationship to Sullivan v Gordon-type damages is not of the same kind as that of the Commissioner of Taxation's relationship to income tax issues, the Australian Competition and Consumer Commission's relationship to trade practices issues or a local government council's relationship to rating issues. 2. A decision on an insurance policy may have an effect on numerous insurance policies issued by insurers. But so far, this Court has not adopted a practice of ordering an appellant-insurer to pay the costs of both parties in this Court and in the courts below. From time to time, the Court may make it a condition of the grant of special leave that an insurer pay the costs in this Court and not disturb the costs orders made in the courts below. But a different area is reached when special leave has been granted without such a condition and the insurer succeeds in the appeal. Unless the insurer has been guilty of some misconduct, the usual order for costs is that costs follow the result. 3. When the Court granted the appellants special leave to appeal in this case, it did not require them to undertake that they would pay the costs in this Court irrespective of the result. Nor did it require an undertaking that they would abide by the costs orders made in the Dust Diseases Tribunal and the Court of Appeal. In these circumstances, I can see no principled justification for requiring them to pay both parties' costs in this Court. The most that can be said in favour of such an order is that overruling Sullivan v Gordon may affect the assessment of damages in an unspecified number of cases in which the first appellant or its subsidiaries are defendants. But Sullivan v Gordon claims were not confined to the first appellant or its subsidiaries. They applied to defendants generally. And one may safely guess that a number of insurance companies would have many more claims for this kind of damages than the appellants have or are likely to have. 4. Similarly, I can see no reason why the appellants should not have their costs in the Court of Appeal. The Court of Appeal, acting on the erroneous assumption that Sullivan v Gordon was good law, ordered the appellants to pay Mr Thompson indemnity costs. On any view of what is fair and reasonable, that order should be set aside. And I can see no reason why the costs in the Court of Appeal should not follow the event. Order 1. The appeals must be allowed. The order of the Court of Appeal of New South Wales dismissing the appellants' appeal against the judgment of the Dust Diseases Tribunal of New South Wales should be set aside. In place thereof should be substituted an order that judgment for the plaintiff be entered in the sum of $300,419.49. The respondent should pay the appellants' costs in this Court and in the Court of Appeal of New South Wales. 2. CALLINAN J. I agree with the reasons and conclusions of the Chief Justice, Gummow and Heydon JJ on the substantive issues in these appeals. 3. With respect however I take a different view on the issue of costs. 4. It is true, as the appellants submit, that as they have succeeded here by having the judgment of the Court of Appeal set aside, they have dislodged the foundation for that Court's order for costs. Notwithstanding this, the respondent submits that the appeals to this Court are analogous with an appeal in a tax case by the Commissioner against a taxpayer, in which the Court often orders that the Commissioner pay the taxpayer's costs of the appeal, without disturbing orders for costs made in the latter's favour below. There are however two important points of distinction here. The first is that offers for the commercial resolution of the matter were exchanged below, and the result in the Court of Appeal formed the basis for the special order for indemnity costs in favour of the respondent there. Secondly, the law on the question raised by these appeals was then far from settled: different intermediate appellate courts had taken different views. In this case the tort was committed in South Australia. If the appellants had been sued in a South Australian court, the governing authority would have been Weinert v Schmidt[204] and the result would have gone the other way. It is only because the appellants were sued in the Dust Diseases Tribunal of New South Wales which was able to exercise long arm jurisdiction[205] and which was bound by Sullivan v Gordon[206], that it has been necessary for the appellants to mount the appeals to this Court. 5. The respondent urges that it is the first appellant who has the greater interest in the outcome because of the precedential value of a decision in its favour on appeal: it has many like cases pending. 6. Unfortunately for the respondent the comparison between this and the tax cases is not well made. The respondent did seek, on the application for special leave, special protective orders with respect to costs but the Court did not however, as it sometimes does in tax cases, require an undertaking by the appellants, as a condition of the grant of special leave, that earlier costs orders not be sought to be disturbed, and that the costs of a successful appeal not be dependent upon the result of it. The two other relevant and more important considerations are, first, that the common law, despite the New South Wales decisions in and since Sullivan v Gordon, could not have been regarded as settled, particularly in light of the different decisions in other jurisdictions. The risks therefore of maintaining a continuing entitlement in litigation to the contentious component were obvious. Furthermore, it should have been equally obvious that the approach to this case was an entirely commercial, and conventionally adversarial one, because each of the parties made offers of compromise, with a view to improving that party's position on costs, apparently carefully calculated by reference to the risks that each was facing. Settlements are encouraged by the courts in the public, as well as the parties', interests. The purpose that they are intended to serve is not to be subverted in a particular case simply because one of the parties has miscalculated his prospects. 7. The respondent should therefore have to pay the appellants' costs of both appeals. 8. The orders of the Court should be: 1. Appeals allowed in both matters. 2. Set aside the judgment and orders of the Court of Appeal of New South Wales dated 26 November 2003 and in lieu thereof, order that the appeal be allowed and the judgment be reduced from $465,899.49 to $300,419.49. 3. Set aside order 1 of the orders of the Court of Appeal of New South Wales dated 17 February 2004 and in lieu thereof, order that the respondent pay the appellants' costs of the appeal to that Court. 4. Order that the respondent pay the appellants' costs of the appeals to this Court. 1. Thompson v CSR Ltd (2003) 25 NSWCCR 113. ↑ 2. CSR Ltd v Thompson (2003) 59 NSWLR 77 (Handley, Sheller and Ipp JJA). The matter of costs was dealt with separately in CSR Ltd v Thompson (No 2) [2004] NSWCA 11. ↑ 3. The plaintiff died after the appeal to the Court of Appeal was instituted but before it was determined. The respondent in this Court is, by order of this Court, an administrator ad litem representing the plaintiff's estate. ↑ 4. Sullivan v Gordon (1999) 47 NSWLR 319. ↑ 5. (1977) 139 CLR 161. ↑ 6. Thompson v CSR Ltd (2003) 25 NSWCCR 113 at 123-124 [42]. ↑ 7. Sullivan v Gordon (1999) 47 NSWLR 319. ↑ 8. Thompson v CSR Ltd (2003) 25 NSWCCR 113 at 123 [40]. ↑ 9. CSR Ltd v Thompson (2003) 59 NSWLR 77 at 80 [12]. ↑ 10. (1977) 139 CLR 161. ↑ 11. (1992) 175 CLR 327 at 332-333, 340 and 347. ↑ 12. Common Law (Miscellaneous Actions) Act 1986 (Tas), s 5. ↑ 13. Motor Accidents Compensation Act 1999 (NSW), s 128; Civil Liability Act 2002 (NSW), s 15 (which does not apply to dust diseases litigation: s 3B(1)(b)); WorkCover Queensland Act 1996 (Q), s 315; Civil Liability Act 2003 (Q), s 59; Accident Compensation Act 1985 (Vic), s 134AB(24)(b); Transport Accident Act 1986 (Vic), s 93(10)(c); Wrongs Act 1958 (Vic), s 28IA; Motor Vehicle (Third Party Insurance) Act 1943 (WA), s 3D; Civil Liability Act 2002 (WA), s 12; Personal Injuries (Liabilities and Damages) Act 2003 (NT), s 23. ↑ 14. [1982] 2 NSWLR 26. This was consistent with earlier authority: Pegrem v The Commissioner for Government Transport (1957) 74 WN (NSW) 417 (Street CJ, Owen and Walsh JJ); Simmonds v Hillsdon [1965] NSWR 837 at 839 per Brereton J; Regan v Harper [1971] Qd R 191 at 194-195 per Hoare J. ↑ 15. [1983] WAR 210. ↑ 16. According to the Scottish Law Commission, Damages for Personal Injuries: Report on (1) Admissibility of Claims for Services (2) Admissible Deductions, Scot Law Com No 51, (1978), par 35, to receive the claim "in the light of the existing principles of the law of reparation in Scotland … at first sight seems startling". This is because in Scots law "damages necessarily involve a loss, either actual or prospective": Edgar v Lord Advocate 1965 SC 67 at 71 per Lord President Clyde. Because the claim was unknown to Scots law, legislative change was recommended (pars 34-44) and effected (Administration of Justice Act 1982 (UK), s 9). ↑ 17. [1997] 2 Qd R 310. It had been accepted earlier in Waters v Mussig [1986] 1 Qd R 224 (Kneipp J). ↑ 18. [1981] 1 WLR 120; [1980] 3 All ER 696. See also Lowe v Guise [2002] QB 1369. ↑ 19. Cummings v Canberra Theatre Trust unreported, Full Court of the Federal Court of Australia, 18 June 1980 (Brennan, McGregor and Fisher JJ), discussed and followed in Hodges v Frost (1984) 53 ALR 373 at 384-385 per Kirby J, Gallop and Morling JJ concurring. ↑ 20. (1999) 47 NSWLR 319 at 331-332 [59]. ↑ 21. Easther v Amaca Pty Ltd [2001] WASC 328 (Scott J: the contrary Full Court decision in Maiward v Doyle, although referred to at [107], was not discussed); Thomas v Kula [2001] WASCA 362 (Wallwork J, Roberts-Smith J, Pidgeon AUJ: Maiward v Doyle was not discussed). ↑ 22. Brown v Willington [2001] ACTSC 100 (Crispin J). ↑ 23. Weinert v Schmidt (2002) 84 SASR 307 (Perry and Williams JJ, Gray J dissenting). The majority gave only brief reasons, but for Perry J's approach see Kite v Malycha (1998) 71 SASR 321 at 340-342. ↑ 24. See the authorities summarised in Carter v Anderson (1998) 160 DLR (4th) 464 at 470-475 (Nova Scotia Court of Appeal). ↑ 25. For example, those of Taylor and Ryan JJA in Kroeker v Jansen (1995) 123 DLR (4th) 652 at 663-674 (British Columbia Court of Appeal). ↑ 26. Dobbs, Law of Remedies: Damages – Equity – Restitution, 2nd ed (1993), vol 2 at 365-367. ↑ 27. Civil Liability Act 2003 (Q), s 59(3). ↑ 28. Wrongs Act 1958 (Vic), s 28ID. In addition, in Sullivan v Gordon (1999) 47 NSWLR 319 at 335 [73]-[75] itself the relevant damages were treated as subject to the limits placed on Griffiths v Kerkemeyer damages by the then equivalent to the Motor Accidents Compensation Act 1999 (NSW), s 128, namely s 72 of the Motor Accidents Act 1988 (NSW). If this conclusion was correct, it was unintended, since in 1988 New South Wales law did not recognise Sullivan v Gordon damages. ↑ 29. Civil Law (Wrongs) Act 2002 (ACT), s 100, replacing Law Reform (Miscellaneous Provisions) Act 1955 (ACT), s 33. ↑ 30. Sullivan v Gordon (1999) 47 NSWLR 319 at 331 [57]. ↑ 31. Sullivan v Gordon (1999) 47 NSWLR 319 at 331 [58]-[59]. ↑ 32. Moragne v States Marine Lines Inc 398 US 375 at 403 (1970) per Harlan J (for the Court); Planned Parenthood of Southeastern Pennsylvania v Casey, Governor of Pennsylvania 505 US 833 at 854 (1992) per O'Connor, Kennedy and Souter JJ (for the Court). ↑ 33. Baker v The Queen [1975] AC 774 at 787-789 per Lords Diplock, Simon of Glaisdale and Cross of Chelsea and Sir Thaddeus McCarthy (holding the Court of Appeal for Jamaica not bound by a Privy Council decision in which "the Board were doing no more than assuming for the purpose of disposing of the particular case, and without any further consideration on their own part, that the proposition of law relevant to the issue of fact in dispute between the parties to the appeal had been formulated correctly by counsel for both parties in agreement with one another"). See also National Enterprises Ltd v Racal Communications Ltd [1975] Ch 397 at 405-406 per Russell LJ, 407 per Cairns LJ and 408 per Sir John Pennycuick; Barrs v Bethell [1982] Ch 294 at 308 per Warner J; In re Hetherington, decd [1990] Ch 1 at 10 per Sir Nicolas Browne-Wilkinson VC. ↑ 34. R v Warner (1661) 1 Keb 66 at 67 [83 ER 814 at 815]. ↑ 35. [1976] AC 443 at 478. ↑ 36. For example Kleinwort Benson Ltd v Lincoln City Council [1999] 2 AC 349. ↑ 37. [1982] 2 NSWLR 26 at 28. ↑ 38. [1982] 2 NSWLR 26 at 36. ↑ 39. [1982] 2 NSWLR 26 at 28-29 and 36-37. ↑ 40. Lowe v Guise [2002] QB 1369 at 1385 [38] per Rix LJ. ↑ 41. [1982] 2 NSWLR 26 at 28. ↑ 42. (1999) 47 NSWLR 319 at 331 [58]. ↑ 43. (1999) 47 NSWLR 319 at 331-332 [59]. ↑ 44. (1999) 47 NSWLR 319 at 322 [2] (a). ↑ 45. (1992) 175 CLR 327 at 333 (emphasis added). ↑ 46. (1992) 175 CLR 327 at 340. ↑ 47. (1992) 175 CLR 327 at 347. ↑ 48. (1992) 175 CLR 327 at 333. ↑ 49. (1992) 175 CLR 327 at 334. ↑ 50. (1992) 175 CLR 327 at 337. ↑ 51. (1996) 187 CLR 354 at 360-361 (emphasis added). ↑ 52. (1996) 187 CLR 354 at 368 (emphasis added). ↑ 53. (1996) 187 CLR 354 at 372. ↑ 54. (2000) 201 CLR 321 at 332 [25] per Kirby J. ↑ 55. In Sturch v Willmott [1997] 2 Qd R 310 at 315-317 and 321 Macrossan CJ and Davies JA appear to have taken the correct view that the recovery they permitted could not be supported by the rule in Griffiths v Kerkemeyer; so did Glass JA in Burnicle v Cutelli [1982] 2 NSWLR 26 at 34. ↑ 56. For example Van Gervan v Fenton (1992) 175 CLR 327 at 346 per Deane and Dawson JJ; Grincelis v House (2000) 201 CLR 321 at 332-333 [25]-[28] per Kirby J and 338-343 [50]-[60] per Callinan J. ↑ 57. Graham v Baker (1961) 106 CLR 340 at 347 per Dixon CJ, Kitto and Taylor JJ; Medlin v State Government Insurance Commission (1995) 182 CLR 1 at 5 and 18 per Deane, Dawson, Toohey and Gaudron JJ and McHugh J respectively. ↑ 58. Arthur Robinson (Grafton) Pty Ltd v Carter (1968) 122 CLR 649 at 658 per Barwick CJ. ↑ 59. (1956) 96 CLR 73 at 79. ↑ 60. It was on the basis of that reasoning that in Scotland Griffiths v Kerkemeyer damages were refused: Edgar v Lord Advocate 1965 SC 67 (Lord President Clyde and Lords Carmont and Migdale, Lord Guthrie dissenting). ↑ 61. Blundell v Musgrave (1956) 96 CLR 73 at 79. ↑ 62. Blundell v Musgrave (1956) 96 CLR 73 at 92. ↑ 63. [2002] QB 1369 at 1385 [39]. ↑ 64. [1997] 2 Qd R 310 at 319. ↑ 65. Cassell & Co Ltd v Broome [1972] AC 1027 at 1086. ↑ 66. This was how Rix LJ frankly put it in Lowe v Guise [2002] QB 1369 at 1385 [38]. ↑ 67. For example, Royal Commission on Civil Liability and Compensation for Personal Injury: Report: Volume 1, (1978), Cmnd 7054-I, pars 352-354. ↑ 68. Skelton v Collins (1966) 115 CLR 94 at 129 per Windeyer J. ↑ 69. Carter v Anderson (1998) 160 DLR (4th) 464 at 473 per Roscoe JA for the Nova Scotia Court of Appeal. ↑ 70. Carter v Anderson (1998) 160 DLR (4th) 464 at 473 per Roscoe JA for the Nova Scotia Court of Appeal. ↑ 71. [1997] 2 Qd R 310 at 321. ↑ 72. See also Carter v Anderson (1998) 160 DLR (4th) 464 at 473, where Roscoe J, speaking for the Nova Scotia Court of Appeal, summarised its reasons for following intermediate appellate courts in Saskatchewan, British Columbia, Alberta, Prince Edward Island and Newfoundland in adopting Sullivan v Gordon principles. ↑ 73. [1997] 2 Qd R 310 at 321. ↑ 74. In Sullivan v Gordon (1999) 47 NSWLR 319 at 322 [2] (c) Mason P employed a similar argument, and Beazley JA quoted the passage at 330 [52]. See also Sharman v Evans (1977) 138 CLR 563 at 598 per Murphy J (in a different context); Lowe v Guise [2002] QB 1369 at 1385 [38] per Rix LJ. ↑ 75. Mason P reasoned to the same effect in Sullivan v Gordon (1999) 47 NSWLR 319 at 322 [2] (b). ↑ 76. [1997] 2 Qd R 310 at 322. Beazley JA quoted this passage in Sullivan v Gordon (1999) 47 NSWLR 319 at 330 [52]. ↑ 77. Burnicle v Cutelli [1982] 2 NSWLR 26 at 28 per Reynolds JA. ↑ 78. Weir, "Compensation for Personal Injuries and Death: Recent Proposals for Reform", in The Cambridge-Tilburg Law Lectures: First Series 1978, (1979) 1 at 18. ↑ 79. [1997] 2 Qd R 310 at 322. The analogy apparently exists in the United States: see above at [9]. ↑ 80. See above at [30]-[31]. ↑ 81. Kite v Malycha (1998) 71 SASR 321 at 342 per Perry J. ↑ 82. See Lowe v Guise [2002] QB 1369 at 1380-1381 [26]-[27], 1382 [29]-[30] and 1385 [38] per Rix LJ. The respondent relied on a passage in Carter v Anderson (1998) 160 DLR (4th) 464 at 473 per Roscoe JA for the Nova Scotia Court of Appeal stressing the importance of the work of injured plaintiffs in maintaining "the health and safety of the family". ↑ 83. Damages for Personal Injuries: Report on (1) Admissibility of Claims for Services (2) Admissible Deductions, Scot Law Com No 51, (1978), par 38. ↑ 84. Griffiths v Kerkemeyer (1977) 139 CLR 161 at 177 per Stephen J and 193-194 per Mason J; Kars v Kars (1996) 187 CLR 354 at 371-372 per Toohey, McHugh, Gummow and Kirby JJ. ↑ 85. Toohey v Hollier (1955) 92 CLR 618. ↑ 86. Commissioner for Railways (NSW) v Scott (1959) 102 CLR 392 at 439-464 per Windeyer J. ↑ 87. For example Brownlee v MacMillan [1940] AC 802 at 809-810. ↑ 88. Burnicle v Cutelli [1982] 2 NSWLR 26 at 31 per Glass JA. ↑ 89. Best v Samuel Fox & Co Ltd [1952] AC 716. ↑ 90. Civil Liability Act 1936 (SA), s 65 (formerly Wrongs Act 1936 (SA), s 33, which was in force at the time of the events leading to the plaintiff's injury). ↑ 91. Law Reform Act 1995 (Q), s 13; Civil Liability Act 2003 (Q), s 58. ↑ 92. Law Reform (Marital Consortium) Act 1984 (NSW), s 3; Common Law (Miscellaneous Actions) Act 1986 (Tas), s 3; Law Reform (Miscellaneous Provisions) Act 1941 (WA), s 3; Civil Law (Wrongs) Act 2002 (ACT), s 218; Administration of Justice Act 1982 (UK), s 2(a). ↑ 93. Transport Accident Act 1986 (Vic), s 93(1) and (2); Motor Accidents (Compensation) Act (NT), s 5; Work Health Act (NT), s 52. ↑ 94. Civil Liability Act 1936 (SA), s 68; Civil Law (Wrongs) Act 2002 (ACT), s 210; Administration of Justice Act 1982 (UK), s 2(c)(ii) and (iii). ↑ 95. Administration of Justice Act 1982 (UK), s 2(b) and (c)(i). ↑ 96. Transport Accident Act 1986 (Vic), s 93(1) and (2). ↑ 97. Motor Accidents (Compensation) Act (NT), s 5 and Work Health Act (NT), s 52. ↑ 98. For example Compensation to Relatives Act 1897 (NSW); Civil Liability Act 1936 (SA), Pt 5 (formerly Wrongs Act 1936 (SA), Pt 2) and other legislation following Lord Campbell's Act – the Fatal Accidents Act 1846 (UK). A quasi-exception is found in legislation ensuring the survival of causes of action available to the deceased victim of a tort, eg Law Reform (Miscellaneous Provisions) Act 1944 (NSW), s 2. ↑ 99. Edmonds v Murphy 573 A 2d 853 at 870-871 (Ct Spec App, Md, 1990) was referred to. ↑ 100. Weinert v Schmidt (2002) 84 SASR 307. ↑ 101. For example, Transco plc v Stockport Metropolitan Borough Council [2004] 2 AC 1 at 9 [6] per Lord Bingham of Cornhill. ↑ 102. [1997] 2 Qd R 310. ↑ 103. Birmingham Corporation v West Midland Baptist (Trust) Association (Inc) [1970] AC 874 at 898. ↑ 104. See, generally, Thompson v Australian Capital Television Pty Ltd (1996) 186 CLR 574 at 603-604, 611-615 per Gummow J. ↑ 105. Law Reform (Miscellaneous Provisions) (Amendment) Act (No 2) 1991 (ACT), s 5. ↑ 106. Daly v General Steam Navigation Co Ltd [1981] 1 WLR 120; [1980] 3 All ER 696. ↑ 107. See Cummings v Canberra Theatre Trust unreported, Full Court of the Federal Court of Australia, 18 June 1980 (Brennan, McGregor and Fisher JJ), discussed in Hodges v Frost (1984) 53 ALR 373 at 384-385 per Kirby J, Gallop and Morling JJ concurring. ↑ 108. Burnicle v Cutelli [1982] 2 NSWLR 26; Maiward v Doyle [1983] WAR 210. ↑ 109. Community Law Reform for the Australian Capital Territory: Second Report: Loss of Consortium: Compensation for Loss of Capacity to do Housework, Report No 32, (1986), pars 18-44. ↑ 110. On the other hand, in Brown v Willington [2001] ACTSC 100 at [109] and [117] Crispin J saw the statutory claim as overlapping, but not coextensive with, Sullivan v Gordon. ↑ 111. See [44] above. ↑ 112. See Esso Australia Resources Ltd v Federal Commissioner of Taxation (1999) 201 CLR 49 at 62-63 [24]-[27] per Gleeson CJ, Gaudron and Gummow JJ, discussing the related issue of whether the common law can be developed by analogy to legislative developments. ↑ 113. [1982] 2 NSWLR 26 at 34. Glass JA's judgment was also relied on in, for example, Sullivan v Gordon (1999) 47 NSWLR 319 at 330 [48] itself. ↑ 114. Burnicle v Cutelli [1982] 2 NSWLR 26 at 35. ↑ 115. [1981] 1 WLR 120; [1980] 3 All ER 696. ↑ 116. Damages for Personal Injury: Medical, Nursing and Other Expenses; Collateral Benefits, Law Com No 262, (1999), pars 3.88-3.92. ↑ 117. [2002] QB 1369 at 1390 [57]. ↑ 118. [2002] QB 1369 at 1386 [41] and 1390 [58]. ↑ 119. Unreported, Court of Appeal (Civil Division), 12 March 1999: the key passages are set out in Lowe v Guise [2002] QB 1369 at 1376-1379 [20]-[22]. Rix LJ contended that Swain's case had no ratio: at 1385 [37]. Potter LJ, a party to Swain's case, said that it had not bound the trial judge to find as he did: at 1387 [50]. ↑ 120. [2002] QB 1369 at 1383 [34]. ↑ 121. Royal Commission on Civil Liability and Compensation for Personal Injury: Report: Volume 1, (1978), Cmnd 7054-I, pars 352-354. ↑ 122. Report on Personal Injury Litigation – Assessment of Damages, Law Com No 56, (1973), par 157. ↑ 123. Damages for Personal Injuries: Report on (1) Admissibility of Claims for Services (2) Admissible Deductions, Scot Law Com No 51, (1978), pars 34-44. ↑ 124. Damages for Personal Injuries: Report on (1) Admissibility of Claims for Services (2) Admissible Deductions, Scot Law Com No 51, (1978), pars 21-23 and 29-32. ↑ 125. Royal Commission on Civil Liability and Compensation for Personal Injury: Report: Volume 1, (1978), Cmnd 7054-I, pars 347-349. ↑ 126. Report on Personal Injury Litigation – Assessment of Damages, Law Com No 56, (1973), par 155. ↑ 127. [1994] 2 AC 350. ↑ 128. Griffiths v Kerkemeyer (1977) 139 CLR 161; Kars v Kars (1996) 187 CLR 354. ↑ 129. Damages for Personal Injury: Medical, Nursing and Other Expenses; Collateral Benefits, Law Com No 262, (1999), pars 3.88-3.90. ↑ 130. Damages for Personal Injury: Medical, Nursing and Other Expenses; Collateral Benefits, Law Com No 262, (1999), par 3.92. ↑ 131. [2002] QB 1369 at 1385 [38] per Rix LJ. ↑ 132. Reliance was placed on Lowe v Guise [2002] QB 1369 at 1386 [41] per Rix LJ and 1388 [52] per Potter LJ. For a discussion of the difficulties in setting limits to recovery, see Sullivan v Gordon (1999) 47 NSWLR 319 at 322-324 [2]-[15] per Mason P. ↑ 133. As in the Australian Capital Territory: Civil Law (Wrongs) Act 2002 (ACT), s 100. ↑ 134. The terms are those of Macrossan CJ in Sturch v Willmott [1997] 2 Qd R 310 at 319: he favoured both. ↑ 135. The Scottish legislation permitting recovery turns on a complex definition of "relative": Administration of Justice Act 1982 (UK), s 13(1). The Law Commission, Report on Personal Injury Litigation – Assessment of Damages, Law Com No 56, (1973), par 157 said the class should be dependants under the Fatal Accidents Act 1846 (UK). ↑ 136. Weir, "Compensation for Personal Injuries and Death: Recent Proposals for Reform", in The Cambridge-Tilburg Law Lectures: First Series 1978, (1979) 1 at 18. ↑ 137. See the questions posed by Macrossan CJ in Sturch v Willmott [1997] 2 Qd R 310 at 315 and 317-318. ↑ 138. As in the Australian Capital Territory: Civil Law (Wrongs) Act 2002 (ACT), s 100. ↑ 139. As in Scotland: Administration of Justice Act 1982 (UK), s 9(3)(a). ↑ 140. See the points made, in a Griffiths v Kerkemeyer context, in Van Gervan v Fenton (1992) 175 CLR 327 at 336 per Mason CJ, Toohey and McHugh JJ; Grincelis v House (2000) 201 CLR 321 at 343 [62] per Callinan J. ↑ 141. Kite v Malycha (1998) 71 SASR 321 at 342 per Perry J. ↑ 142. See Graycar, "Compensation for Loss of Capacity to Work in the Home", (1985) 10 Sydney Law Review 528; Kutner, "Damages for Injuries to Family Members: Does Reform Mean Abolition?", (1993) 1 Torts Law Journal 231 at 250-264 and 278-285. ↑ 143. As in Lowe v Guise [2002] QB 1369. ↑ 144. For the Australian Capital Territory, see Australian Law Reform Commission, Community Law Reform for the Australian Capital Territory: Second Report: Loss of Consortium: Compensation for Loss of Capacity to do Housework, Report No 32, (1986), pars 18-44; and the Civil Law (Wrongs) Act 2002 (ACT), s 100. For Scotland, see the report of the Royal Commission on Civil Liability and Compensation for Personal Injury: Report: Volume 1, (1978), Cmnd 7054-I, Ch 12; Law Commission, Report on Personal Injury Litigation – Assessment of Damages, Law Com No 56, (1973), par 157 and Scottish Law Commission, Damages for Personal Injuries: Report on (1) Admissibility of Claims for Services (2) Admissible Deductions, Scot Law Com No 51, (1978), pars 34-44, together with the Administration of Justice Act 1982 (UK), s 9 (discussed in Lowe v Guise [2002] QB 1369 at 1373-1378 [15]-[20]). ↑ 145. That is, those referred to in notes 17 and 19-22 above. ↑ 146. Simmonds v Hillsdon [1965] NSWR 837 at 839 per Brereton J; Burnicle v Cutelli [1982] 2 NSWLR 26 at 28 per Reynolds JA. ↑ 147. (1999) 47 NSWLR 319 at 329-330 [46] and 331 [57] per Beazley JA. ↑ 148. [2002] QB 1369 at 1380-1381 [26]-[27] per Rix LJ. ↑ 149. (1998) 160 DLR (4th) 464 at 473 per Roscoe JA. ↑ 150. CSR Ltd v Thompson (2003) 59 NSWLR 77 at 79 [7]. ↑ 151. (1977) 139 CLR 161. ↑ 152. Van Gervan v Fenton (1992) 175 CLR 327. ↑ 153. (1999) 47 NSWLR 319. ↑ 154. [1997] 2 Qd R 310. ↑ 155. [1997] 2 Qd R 310 at 316. ↑ 156. Easther v Amaca Pty Ltd [2001] WASC 328; Brown v Willington [2001] ACTSC 100; cf Weinert v Schmidt (2002) 84 SASR 307. ↑ 157. (1999) 47 NSWLR 319. ↑ 158. (1977) 139 CLR 161. ↑ 159. (1961) 105 CLR 549 at 558-559. ↑ 160. Van Gervan v Fenton (1992) 175 CLR 327 at 342, where Deane and Dawson JJ said: "As Stephen J pointed out in Griffiths v Kerkemeyer, however, the distinction between special and general damages has little conceptual relevance to torts such as negligence where identified special damage is not a prerequisite of the cause of action." ↑ 161. (1977) 139 CLR 161 at 193. ↑ 162. Holmes, The Common Law, (1881) at 1. ↑ 163. (1968) 122 CLR 649 at 660-661. ↑ 164. (1968) 122 CLR 649 at 660. ↑ 165. (1980) 65 Iowa Law Review 1249 at 1251. This essay reproduces the text of Professor Atiyah's inaugural lecture delivered before the University of Oxford on 17 February 1978. ↑ 166. Stone, "From Principles to Principles", (1981) 97 Law Quarterly Review 224 at 227. ↑ 167. (1980) 65 Iowa Law Review 1249 at 1251. ↑ 168. Stone, "From Principles to Principles", (1981) 97 Law Quarterly Review 224. ↑ 169. Posner, The Problems of Jurisprudence, (1990) at 454-469; Overcoming Law, (1995) at 11-15, 387-405. ↑ 170. [1964] AC 465 at 516. ↑ 171. Hay v Hughes [1975] QB 790 at 807. ↑ 172. Allen v Waters & Co [1935] 1 KB 200; Schneider v Eisovitch [1960] 2 QB 430; Cunningham v Harrison [1973] QB 942. ↑ 173. Cunningham v Harrison [1973] QB 942 at 952. ↑ 174. Davies v Borough of Tenby [1974] 2 Lloyd's Rep 469 at 479. ↑ 175. (1977) 139 CLR 161 at 170-171. ↑ 176. For example, Grincelis v House (2000) 201 CLR 321 at 330 [19]; Hodges v Frost (1984) 53 ALR 373 at 378. ↑ 177. Kars v Kars (1996) 187 CLR 354 at 372. ↑ 178. Blundell v Musgrave (1956) 96 CLR 73. ↑ 179. (1977) 139 CLR 161 at 179 (emphasis added). ↑ 180. Sullivan v Gordon (1999) 47 NSWLR 319. ↑ 181. Randall v Dul (1994) 13 WAR 205. ↑ 182. Sturch v Willmott [1997] 2 Qd R 310. ↑ 183. (1930) 43 CLR 493. ↑ 184. [1952] AC 716. ↑ 185. [1952] AC 716 at 728. ↑ 186. Law Reform (Miscellaneous Provisions) Act 1941 (WA), s 3; Law Reform (Marital Consortium) Act 1984 (NSW), s 3; Common Law (Miscellaneous Provisions) Act 1986 (Tas), s 3; Civil Law (Wrongs) Act 2002 (ACT), s 218. ↑ 187. Transport Accident Act 1986 (Vic), s 93; Motor Accidents (Compensation) Act (NT), s 5; Work Health Act (NT), s 52. ↑ 188. Attorney-General (NSW) v Perpetual Trustee Co (Ltd) (1955) 92 CLR 113. ↑ 189. [1997] 2 Qd R 310 at 318. ↑ 190. [1997] 2 Qd R 310 at 318. ↑ 191. (1999) 47 NSWLR 319 at 324 [15]. ↑ 192. See reasons of Beazley JA in Sullivan v Gordon (1999) 47 NSWLR 319 at 332 [61]. ↑ 193. [1982] 2 NSWLR 26. ↑ 194. [1982] 2 NSWLR 26 at 34-35. ↑ 195. (1990) 169 CLR 245 at 262. ↑ 196. (1992) 175 CLR 327 at 333. ↑ 197. [1974] QB 454 at 462. ↑ 198. (1968) 122 CLR 649 at 658. ↑ 199. For example, Graycar, "Compensation for Loss of Capacity to Work in the Home", (1985) 10 Sydney Law Review 528. ↑ 200. Graycar, "Compensation for Loss of Capacity to Work in the Home", (1985) 10 Sydney Law Review 528 at 553. ↑ 201. Graycar, "Compensation for Loss of Capacity to Work in the Home", (1985) 10 Sydney Law Review 528 at 549. ↑ 202. In Graham v Baker (1961) 106 CLR 340, this Court held that damages are awarded not merely for diminished earning capacity but because that diminution is or may be productive of "financial loss". In Griffiths v Kerkemeyer the Court reconciled this principle with an award of damages for gratuitously provided services by determining, as a matter of policy, that the wrongdoer should not benefit from the fact that care was provided by family members and others free of charge. ↑ 203. (1984) 53 ALR 373 at 388. ↑ 204. (2002) 84 SASR 307. ↑ 205. Contrast BHP Billiton Ltd v Schultz (2004) 79 ALJR 348; 211 ALR 523. ↑ 206. (1999) 47 NSWLR 319. ↑