De Sales v Ingrilli

GLEESON CJ. In assessing damages to be awarded to a surviving spouse under fatal accidents legislation, what, if any, account should be taken of the chance of the surviving spouse remarrying and thereby obtaining some financial benefit which offsets or diminishes the claimed loss?

The appellant's husband was killed in an accident on the property of the respondent, in Western Australia, on 12 August 1990. The respondent was found to have been negligent, and the deceased to have been partly to blame. Issues of liability are not presently relevant. The deceased was 31 at the time of his death; the appellant 27. The deceased and the appellant were married in April 1985, when they were 26 and 22. There were two children of the marriage, born in October 1987 and April 1990. From 1985 to 1989, the appellant had worked for various periods part-time as a data entry operator. She had also undertaken some part-time study. After her second child was born, she worked full time, without salary, in the family home. The deceased was a practising accountant and tax agent, and was employed as a financial controller and company secretary. He also did some gardening and other domestic work in the home. Following her husband's death, the appellant took up full-time employment as a secretary, office worker and later, stock controller. The appellant had not remarried at the time of trial in 1999. In the period since her husband's death, she has been involved in one relationship of limited duration, in which marriage was never contemplated.

The appellant commenced proceedings in the District Court of Western Australia under the Fatal Accidents Act 1959 (WA) ("the Fatal Accidents Act "). She claimed damages under that Act on behalf of herself and her two children for injury they sustained as a result of the death . Liability was tried first. The respondent was found liable, but his liability was reduced by one-third on account of the contributory negligence of the deceased. The assessment of damages was then undertaken by HH Jackson DCJ. The nature of the claim made on behalf of the appellant and her children will be examined below. The trial judge made no deduction for the general "vicissitudes of life", but applied a discount of five per cent to the appellant's damages to reflect the chance of the appellant obtaining financial benefit from remarriage ("the remarriage discount").

There was an appeal, and cross-appeal, to the Full Court of the Supreme Court of Western Australia. One of the grounds of the appeal and of the cross-appeal concerned the remarriage discount. The respondent argued that the trial judge erred in not applying a significantly higher discount, of say 25 per cent. The appellant argued that no remarriage discount should have been applied. By majority, the Full Court allowed the appeal . Miller J, with whom Parker J agreed, found that the discount applied by the trial judge was "very slight", and that "for a woman of the appellant's age and credentials a 20 per cent deduction would be appropriate." The overall deduction made in relation to the appellant's damages was 20 per cent for the possibility of remarriage and five per cent for general contingencies. The general contingencies applied also in the case of the children. Wallwork J, in dissent, would have rejected this aspect of the appeal on the basis that the trial judge was entitled to apply a discount of five per cent for the contingency of remarriage, even though he himself would have set the discount higher. He did not disagree with the trial judge's decision not to make a deduction for general contingencies, but the brief reasons he gave, and an accompanying reference to a text book, suggests that his Honour may have been addressing a question different from that which the majority had in mind.

The appellant appeals to this Court against the findings of the Full Court as to the appropriate discount to be applied for the chance of the appellant obtaining support from remarriage. The appellant contends that the Full Court and trial judge erred in that they should not have applied any discount for the prospects of remarriage. Alternatively, the appellant submits that the Full Court erred in increasing the remarriage discount. There is also a complaint about the discount of five per cent for general contingencies. The nature of the cause of action

The issue in the present appeal must be considered in the light of the nature of the cause of action being pursued by the appellant, and of the manner in which she formulated her claim for damages.

In Western Australia, as in other Australian jurisdictions, legislation (the Fatal Accidents Act ), provides that, where the death of a person is caused by a wrongful act, and the victim, had he or she survived, would have been entitled to sue, the wrongdoer is liable to be sued in an action brought for the benefit of relatives of the deceased. Such legislation originated in the United Kingdom in 1846 with Lord Campbell's Act . Section 4 of the Fatal Accidents Act provides: "Where the death of a person is caused by a wrongful act, neglect or default, and the act, neglect or default is such as would (if death had not ensued) have entitled the party injured to maintain an action and recover damages in respect thereof, the person who would have been liable if death had not ensued is liable to an action for damages, notwithstanding the death of the person injured, and although the death was caused under such circumstances as amount in law to a crime."

Section 6(1) provides that an action under the Act is brought for the benefit of relatives of the deceased. The term "relative" is defined in Sch 2. It includes, amongst others, a husband or wife, father or mother, son or daughter, people in certain other specified degrees of blood relationship, and any person who, although not married to the deceased, lived with the deceased as husband or wife on a permanent and bona fide domestic basis in certain circumstances and for a certain time. (For the purposes of these reasons it is unnecessary to distinguish between legal and de facto marriage.)

Section 6(2) provides: "In every action [under this Act] the court may give such damages as it thinks proportioned to the injury resulting from the death to the parties respectively for whom and for whose benefit the action is brought." Section 6(4) states that any damages recovered shall be divided amongst the persons for whose benefit the action was brought in such shares as the court sees fit.

Such legislation performs a clear, but limited, social purpose. It seeks to provide some compensation to certain classes of person who suffer in a certain manner in consequence of the death of another, where the death resulted from an actionable wrong. The persons for whose benefit such an action may be brought do not include all those who might suffer injury in such a case; but they are the only persons who can recover damages under the legislation. A person might suffer financial loss in consequence of the death of an employer, a benefactor, a business or professional associate, or someone who stood in some other relationship to that person, including a supportive relationship not covered by the definition of "relative"; but that does not give rise to an entitlement to claim. It is only persons who fall within the defined class of relative for whose benefit an action may be brought.

The Fatal Accidents Act , like its British predecessor, is directed to compensation for "injury". Injury is not defined. It has been interpreted to mean the loss of a benefit the claimant would otherwise have reasonably expected to receive from the deceased, had the accident not occurred. As explained by Windeyer J in Parker v The Commonwealth , two points should be noted about what damages are recoverable for injury. First, damages are calculated by reference to the pecuniary benefit that could reasonably have been expected from the continuance of the life had death not occurred. Damages do not compensate for non-pecuniary injuries such as grief. The provision for apportionment of damages according to "shares" supports the construction of injury as a pecuniary concept . Second, damages for injury are calculated on a balance of pecuniary gains and losses consequent upon the death . In some circumstances, this may mean no damages are recoverable. For example, if a deceased person had earned all of his or her income from capital, and, upon death, that capital was inherited by the surviving spouse, the surviving spouse would have suffered no pecuniary loss .

Historically, the paradigm case under the Fatal Accidents Act was a claim by a dependent wife for damages arising from the death of her husband, who was the family breadwinner. The injury for which damages were recoverable was often described as a loss arising from "dependency". This was the description given to the appellant's claim by the trial judge and the Full Court in this case. In fact, it was the way her claim was expressed in her Statement of Claim. And it was apt in her case. However, injury can occur in circumstances in which there is no dependency. For example, it is now common for both parties to a legal or de facto marriage to have salaried or income-producing occupations. Each may expect to obtain financial advantage from the other, even where they are both fully able to support themselves from their own income, and are therefore not "dependent" in any sense. Characterising the loss as arising from dependency was reasonable in this particular case, but it would be inaccurate and misleading as a comprehensive description of the basis of claims under the Fatal Accidents Act . It would also be erroneous to assume that injury of the kind for which the legislation compensates can only be offset or diminished by a new relationship if that relationship involves dependency.

Further, loss of an expected benefit is not restricted to loss of direct financial support. A claimant's loss may include the value of services the deceased would have provided around the home . A starting point for determining the pecuniary value of these services may be the commercial rate for the provision of the services. In this case, in addition to the loss of a share of the deceased's income, the appellant was awarded modest damages for loss of handyman and childcare services provided by the deceased . Similarly, a husband whose wife worked full-time in the home might recover for the financial loss he suffers as a result of his wife's death, because her services were of value to him. Such an amount is recoverable even if the services are subsequently performed by the surviving spouse or a third party at no cost . With an ageing population, the value of the care provided by one spouse to another may be of increasing importance; and it may be costly to replace. Perhaps a time will come when the paradigm case of a claim under the Fatal Accidents Act will be, not that of a young person injured by the loss of a salary-earning spouse, but that of a person of mature age injured by the loss of a carer. Calculating damages for injury

Calculating damages for the loss of a reasonable expectation of pecuniary benefit usually involves calculating a primary sum and then making such further adjustments or allowances as are necessary to produce a result that gives a true reflex of the loss. The nature of such adjustments and allowances will be influenced by the manner in which the primary sum is calculated. In a case like the present, there are three main elements in determining the primary sum. Each element involves speculative judgments, which cannot be made with accuracy. The court assesses what benefits the deceased would have brought to the family, in the form of either income or the provision of services. The court determines the share of that benefit that would have been enjoyed by a relative during the deceased's lifetime. And the court determines the period for which a relative could reasonably have expected to receive the benefit. For example, a surviving spouse may say that it was reasonable to expect to receive a benefit measured as a share of the deceased's income until the deceased's expected age of retirement. A child of the deceased may reasonably expect to receive such a benefit until the child reaches an age of expected financial independence. The primary sum awarded is the present value of a relative's total expected benefit. The calculation of the primary sum might itself be done by a method that involves allowing for contingencies such as are taken up in actuarial calculations of life expectancy, and the present value of a future income stream.

The court may then be required to allow for further contingencies that may affect the loss of benefit sustained by the claimant. Courts take account of such contingencies in two ways. Certain contingencies may be provided for by way of a general allowance for the "vicissitudes of life". Such contingencies may be relatively unlikely to occur, or their occurrence may be impossible to predict with any accuracy. Other contingencies may be more likely to occur, and more susceptible to specific calculation in the circumstances of a particular case. In these circumstances, if the tribunal assessing damages is a judge sitting without a jury, it may be appropriate to apply a special discount for the specific contingency in question. For example, a general discount is sometimes applied to allow for contingencies such as the chance of premature death, injury, sickness or unemployment. The chance that a person will die prematurely is generally low and is impossible to predict with any accuracy in most cases. However, in some cases it may be clear on the evidence that a particular person has a higher chance of early death, because of an existing illness. In these circumstances, it may be appropriate to apply a larger and separate discount for the specific contingency of premature death.

An example of a case in which a large, and separate, discount was made for particular contingencies is the decision of the Court of Appeal of New South Wales in McIntosh v Williams . That was a widow's claim under the Compensation to Relatives Act 1897 (NSW). The evidence showed that the marriage was very likely to fail. The deceased had a long-standing relationship with another woman, with whom he had a child. The Court of Appeal addressed the contingencies of divorce and remarriage separately from general vicissitudes, and made a discount of 50 per cent on account of those matters.

A typical process of reasoning in relation to a widow's claim is seen in the judgment of the Privy Council in a Canadian case, Nance v British Columbia Electric Railway Co Ltd : "A proper approach to these questions is, in their Lordships' view, one which takes into account and gives due weight to the following factors; the evaluation of some, indeed most, of them can, at best, be but roughly calculated. Under the first head - indeed, for the purposes of both heads - it is necessary first to estimate what was the deceased man's expectation of life if he had not been killed when he was; (let this be ' x ' years) and next what sums during these x years he would probably have applied to the support of his wife. In fixing x , regard must be had not only to his age and bodily health, but to the possibility of a premature determination of his life by a later accident. In estimating future provision for his wife, the amounts he usually applied in this way before his death are obviously relevant, and often the best evidence available; though not conclusive, since if he had survived, his means might have expanded or shrunk, and his liberality might have grown or wilted. ... Supposing, by this method, an estimated annual sum of $ y is arrived at as the sum which would have been applied for the benefit of the plaintiff for x more years, the sum to be awarded is not simply $ y multiplied by x , because that sum is a sum spread over a period of years and must be discounted so as to arrive at its equivalent in the form of a lump sum payable at his death as damages. Then a deduction must further be made for the benefit accruing to the widow from the acceleration of her interest in his estate on his death intestate ... and a further allowance must be made for a possibility which might have been realized if he had not been killed but had embarked on his allotted span of x years, namely, the possibility that the wife might have died before he did. And there is a further possibility to be allowed for - though in most cases it is incapable of evaluation - namely, the possibility that, in the events which have actually happened, the widow might remarry, in circumstances which would improve her financial position."

An example of a similar process of reasoning, in a case where a widower was claiming for the loss of the domestic services of his wife, is seen in the judgment of Deane J in Nguyen v Nguyen : "As has been seen, it is settled that the 'injury' or 'pecuniary loss' for which damages can be recovered is net loss, on a balance of losses and gains. Commonly, in a modern marital relationship in this country, the spouses share, to a greater or lesser extent, the necessary domestic chores and responsibilities. When one spouse dies, the assessment of the value of the lost benefit of the gratuitous services of the deceased as spouse or homemaker must take account of the fact that those services were, at least in part, for the benefit of the deceased as well as for the benefit of the surviving spouse and of the fact that the surviving spouse is relieved of the burden of rendering gratuitous services for the deceased. The injury constituted by the loss may be reduced by the prospect of remarriage."

In both judgments reference is made to the prospect of remarriage as a potential discounting factor. The first judgment was written in 1951; the second in 1990. The appellant's claim

The present appeal is concerned with only two aspects of the manner in which the damages awarded to the appellant were calculated: general contingencies and the possibility of remarriage. Even so, in considering how these subjects were treated, it is essential to bear in mind the way she put her claim. Inevitably, this set the context in which the subject of contingencies arose for decision.

The appellant said that at the time of her husband's death she and her children were financially totally dependent upon the income he brought in to the household from his salaried employment. She contended that, if he had not died, she would have continued to receive the same level of support from him until he reached 65 and perhaps for even longer; each of the children would have been dependent on him until they respectively reached the age of 22. The claim then assessed the deceased's probable earnings from the time of his death until the age of 65, at which he would have retired. It also assessed the extent to which the deceased would have provided for his wife and children out of his income. On the basis of those projections, a loss was calculated and its net present value arrived at. There was also a claim for the value of gardening and domestic services provided by the deceased.

It is obvious that such a claim required consideration of numerous contingencies; some positive, some negative. The assumptions that the marriage would have endured, that the deceased would have remained in his existing employment until the age of 65, that he would then have retired, that the appellant would herself have lived until her husband turned 65, that she would have received the same level of support from her husband over the whole of that period, and that the children would remain dependent until the age of 22, are all self-evidently speculative. But that is the context in which the issue of the possibility of remarriage arose. A widow's claim was made upon the basis that she was fully supported financially by her husband, and could reasonably have expected to continue to receive the same level of support until he turned 65. She also claimed that he provided domestic services of value, for the loss of which she should be compensated. Her claim was for loss of financial support and loss of services. That is the background against which the Western Australian courts allowed for the possibility that such loss might be diminished in consequence of remarriage.

Counsel for the appellant urged on this Court that changes in the role and status of women have made even the consideration of such a possibility outmoded and irrelevant. It may be acknowledged that, in today's society, it is easy to think of individual cases, or of circumstances, in which a widow would not be better off financially as a result of remarrying. It is just as easy to think of individual cases, or of circumstances, in which a married woman would not suffer financial harm as a result of the death of her husband. What follows from that? Some of the arguments of counsel, if taken to their logical conclusion, might suggest that the whole idea underlying the cause of action is now out of date, and that the real solution to the problem is to repeal the Fatal Accidents Act . But we must take the Act, and the appellant's claim, as we find them.

The consequence of making no allowance for the contingency of remarriage (either as part of a general allowance for vicissitudes or as a specific allowance) must be to increase (by a factor of five to 20 per cent, on the approach of the Western Australian courts in this case) the amount to which the appellant is entitled. The primary argument of the appellant, if correct, means that, by reason of changes in the role and status of women, and their increasing independence, a modern widow will be taken to have suffered a significantly greater (not lesser) financial loss in consequence of the death of a husband than her counterpart in earlier times. Remarriage: actual; prospective; or possible

If a claimant under the Fatal Accidents Act has remarried before the date of trial, that remarriage is taken into account because, as Mahoney JA pointed out in A A Tegel Pty Ltd v Madden , it may falsify, or affect, an assumption which would otherwise be adopted in assessing the extent of the injury to the claimant resulting from the death of the deceased. In that case, the Court of Appeal of New South Wales held that a de facto relationship entered into after death and subsisting at trial should be taken into account to reduce the amount to which the claimant was entitled. But even in the case of actual remarriage, what is being allowed for is only a possibility: the possibility that the remarriage will provide pecuniary benefit to the claimant, and that this will wholly or partly reduce the injury sustained by death.

The courts have never assumed that remarriage will inevitably result in such a benefit. In 1968, in Goodburn v Thomas Cotton Ltd Edmund Davies LJ, speaking of possible remarriage by a widow, said: "She may marry a shirker, or a man with ... extravagant personal tastes, or perhaps a man who subsequently walks out on her." There are probably just as many work-shy, or extravagant, or unreliable men now as there were in 1968. But changing social conditions may also have made it less safe to assume that remarriage will be to the financial benefit of a widow or widower. A widow who remarries might, through her own income, support her new husband. A widower who remarries might marry someone who is unwilling or unable to provide the same domestic services as his previous wife. Even so, it is important to bear in mind, as noted earlier, that financial benefit from remarriage does not necessarily involve dependency.

If there has been a remarriage, or if there is in prospect a marriage to a particular person, the court will be in a position to examine the circumstances of the particular case, and these may, or may not, call for a separate, and perhaps significant, discount in a proper case.

Where there has been no remarriage, and no particular marriage is in prospect, there is a double contingency involved: (1) the likelihood of a claimant's remarriage; (2) the likelihood of pecuniary benefit from that remarriage.

In assessing the first question, of the likelihood of a claimant's remarriage, courts have sometimes looked to statistical evidence as to the probability of a widow's remarriage. However, such statistics may not take into account factors such as when the remarriage is likely to occur, and its likely duration. The statistics referred to in argument in the present case are outdated, they are only available for widows, and they relate solely to legal remarriage. Attempts by courts to make a subjective assessment of a particular claimant's chances of remarriage are also fraught with danger. In most cases, courts cannot safely predict, either from statistics or a subjective assessment of the claimant in court, whether the claimant is more or less likely than any other person to remarry.

The second contingency the court must assess is the likely pecuniary benefit from any remarriage. It may be reasonable to assume that, as a general rule, marriage brings certain benefits of pecuniary value, in the form of financial support or assistance, or services. But the court is usually unable to predict what will occur in a particular case. It would be impossible to calculate the actual likelihood of financial benefit by reference to any available statistics. A subjective assessment of the particular probability would again be dangerous, as there is no basis on which the court can predict whether a particular plaintiff will marry "well".

The fact that these contingencies cannot be predicted with any certainty does not relieve the court of the task of taking account of them. There are many uncertainties that attend the contingency of a financially beneficial remarriage: when it occurs, whether it will last and for how long, and whether it is or continues to be financially advantageous. In predicting whether a plaintiff will benefit from a de facto relationship, there may be additional uncertainty. For example, in some states there is no legislative provision for courts to make orders for the division of property between de facto partners, or for the payment of maintenance, upon termination of a de facto relationship . However, these uncertainties are no greater than many that attend the assessment of other "vicissitudes of life", such as the chance of a person becoming unemployed. In assessing the contingency of unemployment, there are uncertainties as to when a person would have lost employment, whether he or she would have been able to find other employment, and if so, when and on what terms. As a minority of the South Australian Law Reform Committee stated : "As far as [the] difficulty [of assessing the contingency of remarriage] is concerned, it seems to us to be of the same character as a great many other conjectural questions which a judge must answer before he can arrive at a just solution to a claim, and we can see no ground in principle or in policy for singling out the factor of remarriage for special exemption."

Subject to the procedural difficulty referred to below, the possibility of a plaintiff remarrying to pecuniary advantage should ordinarily be treated as one of the "vicissitudes of life". Allowance is to be made for the contingency of a financially beneficial remarriage, in the same way as allowance is made for the contingency of premature death, injury, unemployment or financial ruin. It is a chance which usually cannot be predicted with any degree of certainty in a particular case, but which, in the population as a whole, is not a chance that can be disregarded as insignificant.

However, there may be some cases in which there are special or unusual circumstances which make it possible to predict, with some greater degree of certainty, the likelihood of a financially beneficial remarriage. In some cases, a plaintiff may be able to show unusual circumstances which suggest that there is almost no chance of remarriage. Or, it may be arguable that actual remarriage, to a person who offers no financial benefit, effectively precludes the chance of a financially beneficial remarriage . In other cases, a defendant may be able to show special circumstances which suggest that the chances of the plaintiff's loss being reduced by a financially beneficial remarriage are notably higher. Such circumstances include where a person has actually remarried, to his or her pecuniary advantage, before the trial. In these circumstances, there may be concrete evidence which suggests that part or all of the plaintiff's loss will be replaced by benefits received from their new spouse. Similarly, there may be special circumstances where a person is engaged to be married, or living in a de facto relationship, and that relationship is or will be financially beneficial. In such circumstances, the evidence may be less strong than in the case of actual remarriage, but may still be sufficiently concrete to allow a special discount to be made.

Courts have, in some cases, cited a plaintiff's attractive physical appearance, or pleasant demeanour in the witness box, as meriting a higher discount for the possibility of remarriage. However, there is no sound basis for assuming that factors such as appearance, education or job prospects will affect a particular person's chance of financially beneficial remarriage in a predictable manner. Concepts of "marriageability" can be dangerously misleading.

Reference was earlier made to the remark by Miller J about the appellant's "credentials". It seems clear that he was not intending to refer to the appellant's physical appearance. This Court was told that the appellant was not in court during argument of the West Australian appeal, and there is no reason to believe that Miller J had ever seen the appellant or had any idea of what she looked like. Whatever else he had in mind, it cannot have been the appellant's looks. Nor should it have been. As to the appellant's age, there was nothing special about that.

The treatment of the chance of receiving support from remarriage as a factor of modest significance, unless there are special or unusual circumstances which indicate an unusually low or high chance of remarriage, is consistent with the approach taken by the courts in relation to divorce. A court may treat the chance that a plaintiff might have become separated or divorced from the deceased as one of the general contingencies covered by the discount for the "vicissitudes of life". Despite the fact that divorce is now a common occurrence in our society, it is difficult to predict with accuracy in any particular case. Only where there is concrete evidence of marital difficulty or estrangement will there be an assessment of the specific likelihood of divorce in a particular case .

In the ordinary case, the contingency of a financially beneficial remarriage should be treated as part of the "vicissitudes of life". Only in special cases will a separate and substantial assessment of a remarriage discount be warranted. This was not such a case.

There was, however, a procedural reason in this case which led the trial judge to deal with the question separately, as a matter of practical necessity. Having reached the view that other general vicissitudes balanced each other out (a matter to which it will be necessary to return), and having apparently also concluded that the contingency of remarriage was not relevant to the claims of the children (a view that was open to him, and has not been challenged in this appeal), then he was obliged to deal with it separately (if he intended to make allowance for it at all) because it affected only one of the three claimants and not the other two. Given the other aspects of his reasoning, this was a proper course.

Assessing the appropriate allowance to be made for contingencies or vicissitudes is a matter of factual judgment. Until recently, in most Australian jurisdictions it would have been a matter for a jury; and the jury would have been invited to bring to bear their common sense and experience of life. When such an exercise is undertaken by a judge sitting without a jury, the reasoning involved in the assessment will be exposed more clearly to appellate scrutiny. But it remains essentially factual.

A jury's verdict would not reveal any allowance for contingencies. A judge may, or may not, decide to make a separate allowance for some particular contingency, including beneficial remarriage. There is a logical problem about an appellate court accepting that a judge may treat the possibility of beneficial remarriage as one of the vicissitudes of life, to be taken into account with other contingencies, and at the same time, declaring that a judge may not give it any weight. I have difficulty in understanding how this Court can decide that the possibility of beneficial marriage may be taken into account as one of the general vicissitudes or contingencies, and at the same time deny to a trial judge the capacity, as a matter of factual judgment in a particular case, to treat it as increasing the allowance for vicissitudes that would otherwise be made. If it ought to be left out of account altogether, that is one thing. If it may be taken into account, that is another. Once that point is reached, the question becomes one of factual judgment in the particular case.

In my view, no error has been shown in the decision of the trial judge as to the allowance to be made for the possibility of remarriage. It was reasonably open to him to conclude that five per cent was an appropriately modest allowance for the possibility of a financially beneficial remarriage. There was no sufficient justification for the Full Court to increase the five per cent to 20 per cent; and, for the same reason, there is no sufficient justification for this Court to reduce it to nil.

As to the matter of other, or general, vicissitudes the position is more complicated. All the members of the Full Court found various errors in the trial judge's assessment of damages, in relation to matters not the subject of the present appeal. On the question of general vicissitudes the majority took one approach; the reasoning of Wallwork J on the point was brief and enigmatic.

The reasoning of the majority was in conformity with the general approach taken in Western Australia and I am not persuaded that it involved any error. Conclusion

I would allow the appeal.

I agree with the orders proposed by Gaudron, Gummow and Hayne JJ although, for the reasons given, I am of the view that the order to replace order 2 made by the Full Court of the Supreme Court of Western Australia should reflect a discount for remarriage of five per cent and a discount for general vicissitudes of five per cent.

GAUDRON, GUMMOW AND HAYNE JJ. The issue in this appeal is whether, in assessing damages to be allowed in an action brought on behalf of a surviving spouse (or de facto spouse) and children under the Fatal Accidents Act 1959 (WA) ("the Fatal Accidents Act ") in respect of the wrongful death of the partner and parent, the prima facie value of what is lost should be reduced for the contingency that the surviving partner will remarry. Ordinarily, no deduction should be made on this account, whether as a separate deduction, or as an item added to the amount otherwise judged to be an appropriate deduction for the vicissitudes of life, and in this case there should have been no deduction on this account.

The appellant, Mrs De Sales, was born in 1963; her late husband was born in 1959. They married in 1985 and there were two children of the marriage, a daughter born in 1987 and a son born in 1990. Mr De Sales was killed in 1990 in an accident in a dam upon the property of the respondent at Karnup in Western Australia.

Mrs De Sales brought an action in the District Court of that State on behalf of herself and the two children in reliance upon the Fatal Accidents Act . However, Mrs De Sales is not the executrix of the will of her husband and no action was instituted by the executor. Nevertheless, no point has been taken respecting the constitution of the action. The issue of liability was tried separately and first. It was resolved eventually when the Full Court of the Supreme Court upheld a decision of the District Court that there was an entitlement to recover two-thirds of damages to be assessed, the deceased having been one-third to blame by way of contributory negligence.

The assessment of damages then came on for hearing in the District Court in 1999. The primary judge (H H Jackson DCJ) sat without a jury. In the course of his reasons, his Honour said: "[I]n my view a modest reduction should be made for the chance of obtaining support from remarriage, and I deduct five per cent. The deduction should only be made from the share of award that is apportioned to [Mrs De Sales]."

An appeal and cross-appeal were taken on various grounds to the Full Court (Wallwork, Parker and Miller JJ) . One ground taken by Mrs De Sales was that the primary judge had erred in making the deduction of five per cent "in the absence of any or adequate evidence (or findings) as to [her] prospects of remarriage, remarriage rates generally or prospects of financial dependency upon remarriage". On the other hand, in the cross-appeal, the respondent contended that the primary judge had erred in making that deduction because the evidence warranted a significantly higher deduction of "(say) 25%".

In the Full Court, Wallwork J dealt with these matters as follows : "At the time of the trial the appellant was 36 years old. She had successfully re-entered the workforce and was independently redeveloping her career. His Honour was aware of the history of the family since the deceased's death. The appellant had not set her mind against the prospect of re-marriage. At the time of the trial the children were aged 12 and nine. In this case, in my opinion, the learned judge was entitled to take 5 per cent from the damages for the contingency of re-marriage. It was a minimal sum and I personally would have set it higher. That is because the appellant was relatively young and very capable with two children who would not take long to reach adulthood. In my opinion this ground of appeal has not been made out. However neither would I interfere with his Honour's judgment as urged by the respondent."

Parker J agreed with the reasons of Miller J. Miller J differed from Wallwork J. Miller J did refer to a passage in the judgment of Burt J in Hermann v Johnston . There, in the course of dealing with the question of the prospect of remarriage of a widower, Burt J had said : "In such a matter as this, and in the absence of any statistical evidence, and in the absence of any evidence bearing upon the expectations or intentions of the particular plaintiff, it may be thought that the task of assessing the chance of remarrying is beyond the reach of human judgment . But such has been held not to be the case." Miller J expressed his conclusion as follows : "Granted that it is difficult to challenge a trial judge's assessment of the prospects of remarriage or the 'revived capacity to remarry', as it is sometimes put, the fact remains that in this case the learned trial judge's deduction for that contingency was very slight. For my own part, I would think that for a woman of the appellant's age and credentials a 20 per cent deduction would be appropriate." His Honour went on to favour only "a small deduction for the general contingencies of life" and to conclude: "I would consider a deduction for general contingencies of 5 per cent to be appropriate, with the result that in relation to the widow's entitlement, there should be an overall deduction of 25 per cent."

In this Court, Mrs De Sales contends, in the alternative, that the Full Court erred in not accepting her submission that there should not have been any discount by the trial judge for her prospect of remarriage and, if that not be accepted, the majority of the Full Court erred in increasing the figure of five per cent to 20 per cent.

Before turning to these submissions, it is convenient to begin with some consideration of the powers and scope of the relevant provisions of the Fatal Accidents Act and with the relevant antecedents, beginning with the British legislation of 1846, generally known as Lord Campbell's Act .

In estimating damages under Lord Campbell's Act, a jury could not take into consideration mental suffering or loss of society and was permitted to award compensation only for "pecuniary loss". That expression did not appear in the statute. However, in delivering the judgment of the Court of Queen's Bench on a motion for a new trial in Blake v The Midland Railway Company , Coleridge J said : "The title of this Act may be some guide to its meaning: and it is 'An Act for Compensating the Families of Persons Killed;' not for solacing their wounded feelings. Reliance was placed upon the first section, which states in what cases the newly given action may be maintained although death has ensued; the argument being that the party injured, if he had recovered, would have been entitled to a solatium, and therefore so shall his representatives on his death. But it will be evident that this Act does not transfer this right of action to his representative, but gives to the representative a totally new right of action, on different principles. Sect 2 enacts that 'in every such action the jury may give such damages as they may think proportioned to the injury resulting from such death to the parties respectively for whom and for whose benefit such action shall be brought.' The measure of damage is not the loss or suffering of the deceased, but the injury resulting from his death to his family. This language seems more appropriate to a loss of which some estimate may be made than to an indefinite sum, independent of all pecuniary estimate, to sooth the feelings; and the division of the amount strongly leads to the same conclusion: 'And the amount so recovered' ' shall be divided amongst the before mentioned parties in such shares as the jury by their verdict shall find and direct.'" (emphasis added) That decision was followed by the Court of Exchequer in Franklin v The South Eastern Railway Company and subsequently by the Supreme Court of the United States when construing a federal statute based upon Lord Campbell's Act .

In outline, and to some extent in their specific terms, the provisions of the Fatal Accidents Act follow Lord Campbell's Act. In particular, s 4 states: "Where the death of a person is caused by a wrongful act, neglect or default, and the act, neglect or default is such as would (if death had not ensued) have entitled the party injured to maintain an action and recover damages in respect thereof, the person who would have been liable if death had not ensued is liable to an action for damages, notwithstanding the death of the person injured, and although the death was caused under such circumstances as amount in law to a crime." Further, sub-ss (1), (2) and (4) of s 6 state: "(1) (a) Every action brought under this Act shall be for the benefit of relatives of the person whose death has been caused in any manner referred to in section four of this Act. (b) The action shall be brought by and in the name of the executor or administrator of the deceased person as the case may be. (c) In this Act - 'relative' has the meaning given in Schedule 2 to this Act [ ] . (2) In every action the court may give such damages as it thinks proportioned to the injury resulting from the death to the parties respectively for whom and for whose benefit the action is brought. ... (4) The amount of damages recovered, after deducting the costs not recovered from the defendant, shall be divided amongst the persons for whose benefit the action was brought in such shares as the court finds and directs." (emphasis added)

Other expressions, found in the cases, such as "contingencies" and "vicissitudes of life" also do not appear in the statute. Speaking of Lord Campbell's Act in Parker v The Commonwealth , Windeyer J observed : "As the learned authors of the last, the seventh, edition of Winfield on Tort observe, 'the Act is remarkably reticent about what is recoverable and the courts have had their hands pretty full in implementing it on this point'. Their endeavours have not produced an altogether simple body of law. But the governing principles, which are now well established, may be stated in two sentences. The first is from the judgment of Pollock CB in Franklin v The South Eastern Railway Company where he said that the damages 'should be calculated in reference to a reasonable expectation of pecuniary benefit, as of right or otherwise, from the continuance of the life' . The other is from the speech of Lord Porter in Davies v Powell Duffryn Associated Collieries Ltd where his Lordship said that the damages 'are given to compensate the recipient on a balance of gains and losses for the injury sustained by the death' ."

It was decided early in the operation of Lord Campbell's Act that the action was defeated by contributory negligence on the part of the deceased . The apportionment legislation which now applies in jurisdictions including Western Australia expressly applies to actions brought under the Fatal Accidents Act . Hence the apportionment of liability respecting the contributory negligence by Mr De Sales.

It was stressed by this Court in Carroll v Purcell that the balance of gains and losses for which compensation is to be paid must be struck by reference to the gains and losses which must result from the death in question. The ability of a widow to go out to work was said in Carroll v Purcell not to be the result of a revived capacity to undertake gainful employment or a gain resulting from the death of her spouse; her ability to work "was always there" . On the other hand, "[t]he death of one spouse inevitably results in a revived capacity in the other to marry" .

The cases do not readily disclose the stage at which remarriage and the potentiality thereof first was cast into the balance of gains and losses spoken of by Lord Porter in Powell Duffryn . Earlier, in New Zealand , the probability or possibility of remarriage had been treated as too remote. In England, Stephen J had spoken of "a bare chance of receiving some very slight pecuniary help [as] really too remote [a] head of damage" . In the United States, where in many jurisdictions with legislation modelled on Lord Campbell's Act evidence touching remarriage is not admitted, even where this has come to pass before trial , one of the grounds relied upon is that to allow such evidence would require speculation by the factfinder .

What is apparent is that, when accepting the practice, this Court, the House of Lords and the Privy Council have emphasised that what is involved is the prospect of the receipt of material benefit, not merely the reacquisition of marital status itself, and the difficulty in evaluation of that prospect of material benefit.

In Powell Duffryn , Lord Wright spoke of : "having due regard to uncertainties, for instance, that the widow might have again married and thus ceased to be dependent, and other like matters of speculation and doubt". Later, in the course of delivering the advice of the Privy Council in Nance v British Columbia Electric Railway Company Ltd , Viscount Simon had observed with respect to the assessment of damages under the relevant British Columbian legislation : "And there is a further possibility to be allowed for - though in most cases it is incapable of evaluation - namely, the possibility that, in the events which have actually happened, the widow might remarry, in circumstances which would improve her financial position."

In the judgment of this Court in Horton v Byrne , the statement was made: "It is established by authority that in cases of this kind the contingency of the widow remarrying must be taken into account . It is needless to say that no formula has been suggested for arriving at the deduction to be made because of that contingency. It has been left as something which should appeal to the good sense of a jury as an argument for moderation. But it is obvious that age, 'encumbrances' and personality are factors to be weighed. There are contingencies which are alienable and may accordingly be expressed without great difficulty in a money sum. But remarriage is not one of them."

Thereafter, it was said in this Court in Carroll v Purcell of the possibility of a widow remarrying: "This, for what it is worth in any particular case, has so long been regarded as having some value in the assessment of damages in fatal accident cases that it is profitless to debate how far the established rule is justified."

That was said in 1961. The essence of Mrs De Sales' case is that changes since that time merit further consideration as to how far that "established rule" is to be justified in terms of principle. Society has changed markedly since Lord Campbell's Act was first enacted. The Fatal Accidents Act , unlike its predecessor, deals not only with surviving spouses but with survivors of de facto relationships. Very great changes occurred during the last half of the twentieth century in the nature and durability of family relationships, in the labour market, and in the expectations that individual members of society have for themselves and about others - economically, socially, domestically, culturally, emotionally. Even if once it were the case, no longer can a court make any assumption about the role that an individual can be expected to play in the family or in the economy. Yet it is assumptions of conformity to some unstated norm which underpin the making of a "discount for remarriage".

To assess the pecuniary loss that the death has caused the relatives, it is necessary to take account of what may have happened in the future had the death not occurred and, as well, to take account of what may happen to the relatives in the future even though the death has occurred. These predictions, about the "vicissitudes of life", are "very much a matter of speculation" . It follows that the pecuniary loss that has resulted from death cannot be calculated with accuracy. The best that can be done is to assess a sum which will, as far as the limits implicit in the task will permit , represent the value of that loss, assessed at the date of judgment.

Some calculations, which will assist in assessing the value of the loss, can be made. The amount of financial benefit being provided by the deceased to the relatives, immediately before death, can be demonstrated. The present value of that stream of income, if it were to continue into the future, can be calculated (although choosing the appropriate discount rate and the length of time for which the stream is to be assumed to continue will affect the result).

How is account then to be taken of life's uncertainties? Had the deceased not died as he or she did, who is to say how long a life the deceased would have led? What would have happened? Would the deceased have continued to earn at the level being earned before death? Or would death, incapacitating illness or financial calamity (in one form or another) have intervened? If it had, would any of the survivors have then contributed to the financial well-being of the family? Even if there were no disaster (physical, financial or other) would the financial contribution made by the surviving spouse or de facto spouse, have changed anyway? And if action is brought on behalf of a surviving spouse, or de facto spouse, can it be assumed that the relationship would have endured? Will the surviving spouse remarry after the date of judgment, or form some continuing relationship which will have some financial consequence for any of those for whose benefit the action is brought? All these, and more, are possibilities that may have to be reflected in any assessment of the present value of the economic loss suffered by all of the relatives as a result of the deceased's death, not just a surviving spouse. Because the assessment requires estimation and judgment rather than calculation, seldom, if ever, will it be right to express the result as if it were correct to the nearest dollar. That falsely asserts a degree of accuracy in the assessment that is impossible. All that can be done is to select a percentage or lump sum to allow for the estimated value of those possibilities which may or may not have eventuated if the deceased had lived and those which may or may not eventuate in the future.

Of course, in doing that, it is necessary to take due account of the fact that what is sought is a sum for damages that will represent, at the date of judgment, the present value of the benefits which would have been received over time. In assessing those damages, it would be wrong to think that reducing the period during which the benefits will be received from (say) 20 years to 15 years would reduce the amount of damages to be awarded by 25 per cent. The reduction thus effected would be much less than 25 per cent. Conversely, and confining attention to the period of future receipts, to make a deduction of (say) five per cent from the amount that is calculated as the present value of benefits to be received over 20 years is not to assume that the period of benefits that is being considered is reduced from 20 years to 19. It would assume that the benefits would cease much earlier than the end of the nineteenth year. And, of course, a deduction from the present value may reflect not only a reduction in the future period that is being taken to account, it may reflect a reduction in the amount that it is expected will be received. Fixing when that reduction occurs affects the calculation of present value. If it is assumed that the reduction will occur early in the period, the effect on present value is large. If it is assumed that it will occur late in the period, the effect is small.

Statistics may throw some light on some of the questions we have mentioned. They may tell their reader what is the average life expectancy of a person of a certain age. They may reveal how frequent is remarriage among people of a certain age. But great care must be exercised in their use. What are the characteristics reflected in the statistics? Are those relevant to the present inquiry? Why can it be assumed that the individual will conform to the average? To apply a statistical average to an individual case assumes that the case has all the characteristics which, blended together, create the statistic.

The judgment of Windeyer J in Parker v The Commonwealth indicates that his Honour was alive to these considerations. Windeyer J observed : "I was told by the actuary who gave evidence that about one-third of the women who become widows at the age of forty remarry at some time. This piece of information seems to me interesting but not very helpful. So much depends upon matters peculiar to the person and her circumstances, on various factors both emotional and material."

The range of possibilities that lie before those for whose benefit a claim is made under legislation modelled on Lord Campbell's Act is very wide. The financial consequences of some may be to the advantage of the surviving relatives, others may be to their disadvantage. Why should one of those possibilities (remarriage, or the formation of some other continuing relationship) be considered separately from all others? To consider it separately assumes that it is a contingency whose likelihood of occurrence can be separately assessed with reasonable accuracy, and that the financial consequences of its occurrence will, more probably than not, tend in one direction (financial advantage) rather than the other.

Both those assumptions may be flawed. Seldom, if ever, will a court be able to make any useful prediction about whether, or when, one human being will form a close emotional attachment with another. Statistics may provide some basis for saying, in some cases, that it is more probable than not that, at some time over (say) the next 20 years a surviving spouse will form a new relationship. The younger the survivor, the more likely may that be to occur. But, in very many cases, statistics will provide little useful guidance about the time by which it is more probable than not that it will occur. Again, perhaps, the younger the survivor, the easier it may be to fix some outer limit to the time by which the probabilities are that it will occur, but even that would be a hazardous prediction. And it is never assisted by fastening upon some superficial characteristics labelled as "appearance", "personality", "credentials" or the like and having the judge or jury base on those characteristics some estimate of "marriageability".

Even if these difficulties of predicting that a surviving spouse will form some new continuing relationship were to be surmounted, the financial consequences of its occurrence are even less predictable. Who is to say that the new relationship will endure, and that, if it endures, it will provide financial advantage to the person who is now the surviving spouse? And if it is a financially beneficial relationship at its outset, who is to say what events will intervene thereafter? Will the new spouse or partner suffer some catastrophe and the person who is now the surviving spouse then have to care and provide for the new partner, the children of the first union, any children brought by the new partner to the new union, and any children born of the new union? Who can say?

It is these last points about the financial consequences of a new relationship which are of critical importance. They deny the validity of looking separately at some "discount for remarriage" over and above whatever discount is made for the "vicissitudes of life". Among those vicissitudes are all the hazards and benefits that may befall a person or, where the claim is made for a surviving spouse, may befall a couple during life. Any new union, which is formed after the termination of the union which underlies a claim made pursuant to a wrongful death statute modelled on Lord Campbell's Act, is as exposed to precisely the same kinds of hazard and danger as was the earlier union. It, too, may end in death, separation or divorce. The financial advantages and disadvantages to one partner will change throughout the continuance of the union as the careers and ambitions of the partners change both with and against their will. Those, who today are receiving income from personal exertion, may, tomorrow, cease doing so for any number of reasons. Those who are employed may have the employment terminated. Those who are self-employed may fall ill, or the venture in which they are engaged may fail. Those who receive income from investments may invest unwisely or unprofitably. Those who are now not employed outside the house may later forge a new career either because they want to or because they feel they should or must do so. And so the examples can be multiplied. Yet if these possibilities are taken to account in assessing the vicissitudes to which the former union was subject (and they must) to ignore them when considering a new union, by assuming that the new union would be destined to survive and prosper, would be to shut one's eyes to reality.

It is, therefore, wrong to treat the prospect of remarriage or the prospect of forming some new continuing relationship as a separate item for which some identified discount must be made from whatever calculation is made of the present value of future benefits that would have flowed from the deceased to the relatives. Even if the prospects that a surviving spouse would remarry or enter a new continuing relationship could be assessed (and there will be few cases where that would be possible), predicting when that would occur is impossible, and predicting some likely outer limit of time by which it would probably have occurred is only slightly less difficult. But most importantly, it cannot be assumed that any new union will be, or will remain, of financial advantage to any of those for whose benefit the action is brought. That being so, some financially advantageous marriage or relationship must be treated as only one of many possible paths that the future may hold. It is wrong to single it out for special and separate allowance. That others in the past have had damages reduced on this account is not reason enough to continue the error.

Nor can the prospect of remarrying or forming a new relationship properly be seen as a matter which, under the general heading of "the vicissitudes of life", enlarges the discount which otherwise must be made from the present value of the benefits which the deceased was providing at death. The assessment of that discount is not easy. It must reflect not only the fact that the future may have been better than the past but also the fact that it may not. It is wrong to fasten upon one of the myriad possible paths that life may take and say that, on account of that possibility, it is right to enlarge the discount that must be made. The discount can be assessed only as a single sum which reflects all of the possibilities.

That is not to say that, if there is evidence at trial that a new relationship has been formed, account may not be taken of evidence revealing whether that brings with it financial advantage or disadvantage. It would be wrong to adopt the rule followed in some American jurisdictions and require the tribunal of fact to assess the damages without that evidence. If the relationship is reflected in marriage, or if there is relevant legislation creating rights between de facto partners, the property rights of the partners will no doubt loom large in that assessment. Likewise, if there is evidence that a surviving spouse (or de facto spouse) intends, at the time of trial, to establish such a relationship with an identified person, account may be taken of evidence of the probable financial consequences of that relationship. In each case, however, it would be wrong to assume that the financial consequences revealed in evidence will inevitably continue.

Ordinarily, such cases apart, no separate allowance should be made for the possibility, even probability, that a new relationship will be formed. That is because it cannot be said, even on the balance of probabilities, whether, having regard to the whole of the period which must be considered, that relationship would be to the financial advantage or disadvantage of those relatives of the deceased for whose benefit the action is brought. If the period to be considered is very long (as would be the case where the surviving relatives are young) the range of possibilities that must be considered is very large. By contrast, if the period is short (where the surviving relatives are older) the possibilities of death, illness, loss of employment and the like are all the higher. No doubt, as was said more than 70 years ago in the United States District Court : "If we should enter upon an inquiry as to the relative merits of the new husband as a provider, coupled with his age, employment, condition of health, and other incidental elements concerning him, unavoidably we should embark upon a realm of speculation and be led into a sea of impossible calculations." But the critical point is not that the inquiry is hard, it is that it is an inquiry which does not lead to a useful answer. In the end, all that can be said is that the future is uncertain. The value of what is lost as a result of the wrongful death must strike a balance of all the gains and losses that have been and may thereafter be suffered. There is no basis for fastening upon some to the exclusion of others.

Although the primary focus of the argument in the appeal was upon the deduction made by the Full Court for the prospect of remarriage, it was also argued that the Full Court erred in allowing a five per cent deduction for general contingencies both for the appellant's and the children's entitlements. The trial judge made no allowance on that account and it was argued for the appellant that, no error having been disclosed in the trial judge's reasoning, the Full Court should not have made the deduction it did.

It will be apparent from what has been written above that, as a general rule, some allowance should be made for the various possibilities that impact upon the assessment of the pecuniary loss suffered in consequence of the wrongful death of a partner and/or parent. Of course, it is not to be assumed that the possibilities are all adverse. As Windeyer J pointed out in Bresatz v Przibilla : "All 'contingencies' are not adverse: all 'vicissitudes' are not harmful. A particular plaintiff might have had prospects or chances of advancement and increasingly remunerative employment. Why count the possible buffets and ignore the rewards of fortune? Each case depends upon its own facts. In some it may seem that the chance of good fortune might have balanced or even outweighed the risk of bad."

Although the trial judge did not say so, it would appear that his Honour took the view that, so far as the deceased was concerned, the chance of the deceased improving his earnings over time balanced or outweighed negative contingencies which had to be taken into account. Certainly, there was evidence which may have led him to form such a view. However, adverse possibilities such as illness, the loss of employment due to economic downturn and, even, early retirement had to be taken into account. Moreover, allowance had to be made for possible future events, including that the children might become financially independent (or partially so) before the age of 22, that being the age selected by the trial judge before which they would not do so.

In Kember v Thackrah , the standard discount for contingencies in Western Australia was said to be "in the vicinity of 2 per cent to 6 per cent". Apart from the possibility that the deceased's earnings may have improved over time, nothing was put in argument to suggest that the present case should be approached on the basis that it has special features warranting departure from the standard or norm which applies in Western Australia.

The fact that the deceased's earnings may have improved over time is but one aspect of a range of possibilities, some fortunate and some unfortunate, that must be taken into account in estimating the extent of the pecuniary loss suffered in consequence of his death. The discount of five per cent was within the standard range and makes appropriate allowance for the various contingencies, including the prospect of the appellant's entering into a permanent relationship which is to her financial benefit. The argument that the Full Court erred in allowing a five per cent discount should be rejected.

The appeal should be allowed to the extent necessary to set aside order 2 of the Orders of the Full Court made on 1 December 2000. The appellant should have her costs of the appeal to this Court, but the costs orders made by the Full Court in relation to the appeal and cross-appeal to that Court (orders 4 and 5) should stand, as should the order allowing the cross-appeal (order 3). The parties should have an opportunity to agree upon the amount in which, in accordance with these reasons, judgment should now be entered (an amount which, it would be expected, would be rounded to the nearest thousand dollars to reflect the imprecision of the calculation). Accordingly, there should be a further order that, if within 28 days of the date of the order of this Court, the parties submit to the Registrar a signed minute of their agreement to the amount for which judgment should be entered, in place of order 2 of the Orders of the Full Court, there should be judgment for that sum. If the parties do not, within that time, submit such a signed minute, there should be an order that the matter is remitted to the Full Court.

McHUGH J. By the law of Western Australia, a person who has suffered the loss of financial support because the defendant negligently caused the death of that person's spouse or de facto spouse is entitled to have the defendant pay compensation for that loss. The issue in this appeal is whether, in assessing that compensation, the defendant is entitled to a discount for the chance that the plaintiff may receive financial support by remarrying or entering into a de facto relationship with another person.

In my opinion, where there is more than a remote chance that the plaintiff is likely to receive financial support from remarrying or entering into a de facto relationship, the defendant is entitled to have the damages payable in such a case discounted by the percentage that reflects that chance. It is logical and it is just that a widow or widower who claims compensation for the future loss of financial support from a deceased person should give credit for any financial support that has replaced the support that that person would have received from the deceased. And it is logical and it is just that the damages awarded should also be discounted to reflect any probability - high or low - that the plaintiff will receive support in the future from remarrying or entering into a de facto relationship. To hold otherwise is to give the plaintiff a windfall - in many cases a substantial windfall - and to require the defendant to pay the plaintiff more than that person has lost financially. I also think that there is no advantage - and some danger - in subsuming the discount for future support under the head of general contingencies. Because the appropriate discount must vary considerably depending on the age of the widow or widower, the general contingencies percentage would have to be varied in each case. If the variation is done properly, subsuming the discount under general contingencies will be of no practical value. Statement of the Case

Teresa Margaret De Sales sued Albert Ingrilli in the Supreme Court of Western Australia for damages for the financial loss that she has suffered as the result of Mr Ingrilli negligently causing her husband's death. Her claim was made under the Fatal Accidents Act 1959 (WA). The Supreme Court awarded her damages but reduced the amount that represented her financial loss by one-third because of her husband's contributory negligence and by five per cent to reflect the chance that she might obtain financial support as the result of remarriage. Mrs De Sales was aged 27 when her husband was killed on 12 August 1990. There were two children of the marriage, one born in October 1987 and the other in April 1990. Both parties appealed to the Full Court of the Supreme Court. By majority the Full Court increased the discount for the chance of remarriage to 20 per cent and deducted a further five per cent for the general contingencies of life.

Following a grant of special leave to appeal, Mrs De Sales appeals to this Court against the order of the Full Court. She contends that, in actions under the Fatal Accidents Act and its analogues, courts should not reduce the damages otherwise payable to a plaintiff because of the chance that the plaintiff may obtain financial support in the future as the result of remarrying or entering into a de facto relationship. Alternatively, she contends that in her case the discount should be reduced to the five per cent awarded by the trial judge in the Supreme Court. The Fatal Accidents Act

Section 4 of the Fatal Accidents Act declares that, where the wrongful act of a person causes the death of another person and the deceased, if he or she had lived, could have maintained an action for damages against the wrongdoer, the wrongdoer "is liable to an action for damages". Section 6(1) declares that the action is brought for the benefit of the relatives of the deceased. They are defined in Sched 2 to include a husband or wife, a son or daughter and a father or mother. They also include any person who, though not married to the deceased, lived with the deceased as husband or wife on a permanent and bona fide domestic basis for a specified period in specified circumstances. Section 6(2) empowers the court to "give such damages as it thinks proportioned to the injury resulting from the death to the parties respectively for whom and for whose benefit the action is brought."

The Fatal Accidents Act is based on Lord Campbell's Act 1846 (UK) . Before Lord Campbell's Act , no action for damages for or arising out of the killing of a person could be maintained in a civil court . That Act gave a right to the relatives to recover damages "proportioned to the injury resulting from the death". But from the beginning the term "injury" was read as confined to pecuniary loss . And Justices of this Court have accepted that that is so . In Davies v Taylor , Lord Reid said that the "injury" "must be of a financial character" and that it meant the "loss of a chance" . That is to say, damages are awarded under Lord Campbell's Act for the chance that the deceased would have provided the relative with financial support or its equivalent in the future. The damages are "for the loss of the expectation of financial support by the deceased" . Thus, the tribunal of fact in assessing damages must value the chance that each relative had of obtaining a financial benefit from the deceased if that person had not been killed by the defendant.

In determining the extent of the "injury", financial gains as well as losses resulting from the death must be taken into account . In Davies v Powell Duffryn Associated Collieries Ltd , Lord Russell of Killowen said : "The general rule which has always prevailed in regard to the assessment of damages under the Fatal Accidents Acts is well settled, namely, that any benefit accruing to a dependant by reason of the relevant death must be taken into account. Under those Acts the balance of loss and gain to a dependant ... must be ascertained, the position of each dependant being considered separately."

Sentiment has no place in the action. The relatives are not compensated for grief, sorrow or bereavement. "It is," said Lord Wright, "a hard matter of pounds, shillings and pence" . So pervading is the notion that "injury" is concerned with the loss of the chance of financial support that the funeral expenses of the deceased were not recoverable by the relatives in an action under the Act until the legislature intervened in England and some Australian States.

In a report on the Fatal Accidents Act , the Western Australia Law Reform Commission has emphasised that the relatives must prove, and can only prove, financial loss or its equivalent in an action under that Act. The Commission said: "In an action under Western Australia's Fatal Accidents Act 1959 , only recovery of economic or material loss is allowed. Where the deceased is a wife and mother, damages can be claimed for loss, not only of outside earnings by which she contributed to the family purse, but also of her domestic services in looking after the home, husband and children. But the husband cannot recover for the loss of her companionship or for the loss of her love. Nor can he recover for grief or mental suffering which he endures because of her death."

In an action under the Fatal Accidents Act , the first issue is to determine the value of the gains that the relatives would have obtained from the deceased if he or she had not died. That inquiry requires the court to determine what would have happened if the deceased had lived, a course that immediately puts the tribunal of fact into the impossible position of having to make assumptions and predictions about a future that cannot occur. Even in simple cases, the probability must be very high that the assumptions and predictions are wrong and wrong by a wide margin. As the Full Court of the Supreme Court of New South Wales pointed out in Phali v Commissioner of Railways : "All but the simplest claims under [Lord Campbell's Act] present uncertain and imponderable elements, so that an accurate arithmetical approach is quite impossible."

In most cases, the starting point of the inquiry will be the income of the deceased at the time of death and how much of that income went to the benefit of the relatives. Unless the income of the deceased was very high, the evidence showing the relatives' benefit at the time of death will probably be determined by taking the deceased's income and deducting an amount to cover the cost of the deceased's food, clothing and personal expenditure. Such evidence may range from that of the surviving spouse painting a picture of the deceased as a frugal, shabbily dressed, selfless provider for the family to more sophisticated evidence, based on Household Expenditure Surveys of the Australian Bureau of Statistics. Once the cost of the deceased's support is deducted, judges and juries almost automatically assume that the relatives have had the benefit of the residue of the net income of the deceased. To this residual sum will be added a sum for any services, measurable in money, which the deceased provided for the family. Thus, there is room for large errors even in the relatively simple task of estimating the financial dependency of the family at the date of death. But the scope for error at this stage is almost insignificant compared to the scope for error in determining the benefits that the family would have received if the deceased had survived.

Assessing future financial benefits depends on the accuracy of assumptions or predictions concerning: . the deceased and his or her spouse meeting the actuarial determination of the period of their joint life expectancy, . the marriage of the deceased and the spouse surviving to the end of their joint life expectancy and not joining the one-third of first marriages or the one-half of second marriages that end in divorce, . the income of the deceased not being cut off by premature death or reduced by sickness, accident, unemployment, retrenchment or demotion, . the deceased continuing to maintain the same or similar level of financial support to the dependants that they were receiving as at the date of death, . the relatives receiving non-income financial benefits from the deceased such as superannuation or an increase in wealth through investments or inheritance.

Allowance must be made for any financial gains that the relatives derive as the result of the deceased's death. The scope for error in making the allowance will depend on the nature of the gain. In practice, the scope for error is usually high because it often involves calculating the difference between what the family now gets and the present value of benefits that the judge or jury assumes would have been available to the relatives in the future.

The younger the surviving spouse was at the date of death, the greater is the scope for error in assessing future financial benefits and present gains. To compensate for these contingencies, judges make, and juries are directed to make, an allowance usually in the form of a percentage deduction from the amount otherwise to be awarded. In Western Australia, the figure for general contingencies is about five per cent in a claim for damages for personal injury and, as the present appeal shows, five per cent in a claim for damages for wrongful death . In New South Wales, the conventional discount for general contingencies is 15 per cent in personal injury claims. Some years ago I was a party to the reasons in a special leave application which stated that there was "no reason to doubt the correctness" of a 15 per cent discount for the vicissitudes of life . Further reflection on statistics concerning unemployment and the payment of social services and workers' compensation, however, now makes me think that the figure of 15 per cent is too high - at least for low to middle income workers. And the Western Australian figure, although probably too low, may just be within the appropriate range in personal injury claims, particularly where the actuarial figures or tables used to calculate the present value of lost earnings allow for death. Beneficial social and industrial legislation has improved, and continues to improve, the financial lot of most workers who lose work and income through injury, illness, accident or unemployment. Once allowance is made for contingencies being favourable as well as adverse, a figure of five per cent may not be unreasonably low for the contingencies involved in personal injury claims. But I do not think the same can be said of a five per cent discount for general contingencies in wrongful death claims. Given the ever threatening spectre of divorce, its high rate, and the other contingencies in a wrongful death action, a figure of five per cent for general contingencies in that class of action seems unreasonably low.

Whatever figure is used for contingencies in a wrongful death claim, however, it would be a miracle if the amount awarded in a particular case got near the correct figure. Fortunately for judges and juries - unlike investment fund managers - their assessments of future earnings and dependencies cannot be proven to be wrong. At all events, they cannot be proven wrong unless the amount awarded bears no reasonable relationship to the figures proved in evidence.

However, the risk of error in assessing the general damages will be increased if the judge or jury does not determine and value the chance of the financial loss to the surviving spouse and relatives being reduced because of financial and other support from remarriage or a de facto relationship. If the surviving spouse remarries before the trial or between the trial verdict and hearing of an appeal , the court assesses the claim on the basis that the survivor will receive financial support that replaces the lost support from the deceased. Similarly, if the surviving spouse has entered into a de facto relationship before the trial or appeal, the court will take any financial support or its equivalent into account . To accept this principle does not mean that the surviving spouse is not entitled to any damages beyond the date of remarriage or entering into the de facto relationship. That may sometimes be the result when the second marriage seems to provide greater support and security than the marriage to the deceased . But if the second marriage is not as secure or supportive as the first marriage, the surviving spouse may still obtain substantial damages. In some cases, the survivor may obtain damages almost as great as if he or she had not remarried or entered into a de facto relationship .

For a very long time - perhaps even from the commencement of these actions in the 19th century - the courts have also assessed and valued the chance that the surviving spouse will obtain financial support or its equivalent from remarrying . In recent times, the courts have also assessed and valued the chance that the surviving spouse will obtain financial support or its equivalent from a de facto relationship .

To determine and value the chance of support from such relationships accords with the basic principle upon which damages in this type of action are determined. "What must be ascertained," said Dixon J "is whether any and what loss has been sustained by the relatives of the deceased after comparing the material benefits depending upon his life with any material gains accruing from his death." Where financial support or its equivalent from remarriage or a de facto relationship replaces, wholly or partly, the support given by the deceased, the "injury" resulting from the death is to that extent reduced.

The appellant contended that, despite the longevity of the practice of valuing the chance of support from remarriage, this Court "should consider abolishing the practice of discounting widows' awards for the prospects of financial dependency upon remarriage." She argued that the discounts varied greatly from judge to judge, that judges had stated that the "exercise is not only unattractive but is not one for which judges are equipped" and that the United Kingdom and Northern Territory legislatures had abolished the practice. None of these arguments warrant this Court abolishing the practice of discounting. Figures supplied by the Australian Bureau of Statistics, based on registered remarriages in 1997, 1998 and 1999, show that 42.2 per cent of widows aged 36 - the appellant's age at time of the trial - will remarry. Those figures also show that 62.2 per cent of widows aged 27 - the appellant's age when her husband died - will remarry. If the courts were to ignore these figures - which disregard de facto relationships - and were to abolish valuing the chance of future financial support from remarriage or a de facto relationship, they would be overcompensating surviving spouses in the great majority of cases. The task of valuing the chance of future support is no doubt difficult. But it is no more difficult than valuing many of the other lost chances that courts are regularly called on to value. It hardly seems more difficult than valuing the loss of the chance of winning prize money in a beauty competition or the loss of the chance of earning betting and prize money from the training and racing of a horse .

The actions of the United Kingdom and the Northern Territory legislatures in abolishing the practice provide no sound reason for abolishing the discount. That legislation is inconsistent with the principles on which courts assess damages in wrongful death cases. Professor Atiyah has strongly criticised the 1971 United Kingdom legislation abolishing the discount. He says that it is irrational. He has written : "This must be one of the most irrational pieces of law 'reform' ever passed by Parliament. It would be as sensible to require a divorced husband to maintain his wife after she has remarried, or for the State to pay widows' pensions after remarriage." He illustrated the absurdity of the legislation by the reported case of a 25 year old widow obtaining very substantial damages for the death of her husband although within two years she had remarried an oil company executive with a substantial salary .

Furthermore, I find it difficult to see why the issue of support from remarriage or a de facto relationship is any more distasteful for a judge or jury than determining many issues in personal injuries action. Assessing damages for mental impairment, scarring or loss of libido or sexual function, for example, is a task that commonly falls for determination in those actions. From time to time in a wrongful death claim, the court may have to determine an issue more unpleasant than the prospect of future support from remarriage or a de facto relationship. That issue is whether the relationship of the parties had reached a state that indicated that their marriage would not have lasted much beyond the date when the deceased died.

In Buckley v John Allen & Ford (Oxford) Ltd , Phillimore J doubted whether judges were "equipped" to decide the remarriage issue. But if judges and juries are competent to determine the joint life expectancy of the couple, the future state of the marriage, the future income of the deceased, the future financial benefits that the surviving spouse and relatives would gain from the deceased, I cannot see why they are not competent to determine and value the chance of financial support from remarriage or a de facto relationship. Moreover, de facto spouses now qualify as dependants under many wrongful death statutes. Determining the dependency and lost benefits of such claimants can hardly be said to be more speculative than determining the chance of support from remarriage or a de facto relationship.

The ready availability of reliable and extensive statistics touching on the issue of future support probably makes the task of determining the prospect of support from remarriage easier now than it has ever been. Reliable statistics are available that show the probable chance of a widow at any age remarrying. Reliable statistics are also available that show the number of people living in de facto relationships. Unless there is some definite reason for concluding that those statistics do not apply to a widow (or widower), they should be regarded as indicating the probability that the particular plaintiff will remarry or enter into a de facto relationship. In the case of the individual plaintiff, using the statistics may be over- or under-inclusive. But overall they will provide a more reliable guide to the chance of remarriage than the tribunal of fact assessing that chance by determining the attractiveness of the plaintiff or accepting the usually sincere claims of widows or widowers that they will never remarry.

Judges are sceptical of using statistics and theories of statistical probability. An extreme example is found in the judgment of Meagher JA in Government Insurance Office of NSW v Rosniak where his Honour said the trial judge had correctly rejected the evidence of the actuaries who were called as witnesses in that case. His Honour said that they had thrown "themselves into the task of forecasting the events of the next sixty-one years like ancient Etruscan soothsayers examining the entrails of sacrificial birds." His Honour's comments would have disappointed Oliver Wendell Holmes, Jnr. Writing over 100 years ago, the great American jurist argued that "[f]or the rational study of the law ... the man of the future is the man of statistics and the master of economics" . The comment of Meagher JA suggests that "the future" has not yet arrived.

Using descriptive or inferential statistics cannot ensure an accurate assessment of damages in any particular case. But in determining future probabilities, there is as of now no better way. In the long run, using descriptive and inferential statistics will prove more accurate in determining wrongful death cases than relying on the intuitions of judges and juries based on their impressions of plaintiffs and their assumptions of what people like the plaintiff are likely to do. No modern society or government could continue to exist in its present form without using statistical data and the conclusions that are reached by applying statistical and probability theory to that data. I see no reason why courts should not invoke the aid of such powerful predictive tools, whenever it is feasible to do so. In this particular area of the law, the search for highly individualised justice borders on delusional.

Determining the probability of the surviving spouse remarrying, however, is only the first step in determining the chance of the plaintiff receiving future financial support that reduces his or her loss. A harder task is to determine the quantum of support that the plaintiff might receive in the future. But difficult as this always has been, and probably always will be, it seems a reasonable assumption that, in the majority of cases, the plaintiff will obtain the kind of support that he or she obtained from the deceased. At all events that assumption seems no less reasonable than the assumption that the deceased would have continued to have the same income stream and the relatives would have had the same level of support from the deceased in the future. Moreover, the defendant has the onus of establishing the benefits that may be obtained from future financial support . If the evidence suggests that the surviving spouse may not receive the same level of support from an existing or future relationship, it is the defendant who must bear the consequences.

Accordingly in my view, this Court should not abolish the long established rule that, in a wrongful death action, the court must assess and value the chance of the surviving spouse obtaining financial support in the future from remarriage. Nor should the Court abolish the more recent rule that support from a de facto relationship is a matter that may be assessed and valued.

Moreover, I see no advantage in subsuming the discount for future support under the rubric of general contingencies. The appropriate discount for support must vary considerably depending on the age of the widow or widower. According to the figures of the Australian Bureau of Statistics to which I have referred, a 30 year old widow has a 55.7 per cent chance of remarrying in her lifetime while a 60 year old widow has a 5.3 per cent chance. Any discount must take account of these widely varying figures. It is impossible to apply the same discount figure to widows of substantially different ages. If the support discount were subsumed under the head of general contingencies, the percentage for general contingencies would have to be adjusted on a case-by-case basis to reflect the different chances of future support. If the variation is done properly, it would move in accordance with the age and circumstances of the widow or widower. If it is done properly, subsuming the discount under general contingencies will be of no practical value. Moreover, there is a danger that to subsume the discount under general contingencies will lead to judges and juries applying substantially similar discounts, irrespective of the age and prospects of the widow or widower in question.

In the present case, neither the trial judge nor the Full Court had the advantage of any statistics concerning remarriage. However, I can see no reason why we should interfere with the assessments of the Full Court concerning the remarriage discount and general contingencies. In an area where minds will reasonably differ as to the proper percentages, I cannot see how - given all the contingencies that pervade this class of claim - it can be said that a total deduction of 25 per cent for this widow is too high. And with the greatest respect to those who think that a five per cent total deduction is appropriate, I think that such a percentage is unreasonably low. It means that, despite all the assumptions and uncertainties in assessing the loss of the chance of dependency over a 34 year period, the figure placed on the value of the chance by the trial judge has a 95 per cent confidence level of being correct.

Moreover, given the appellant's age and circumstances, it would seem to follow from the five per cent discount applied to her claim that no widow or widower will suffer a higher discount than five per cent unless possibly that person has already remarried or entered into what appears to be a permanent de facto relationship. It also suggests that, given the standard discount for contingencies in Western Australia, there has either been no discount for the chance of remarriage or at best no more than one per cent. The five per cent figure must also create a dilemma for courts in New South Wales and the Australian Capital Territory and any other jurisdiction where a 15 per cent figure is generally used. If the appellant's claim is reduced by only five per cent, it would seem difficult for courts in those jurisdictions to impose a higher discount for general contingencies in wrongful death claims in their jurisdiction. That will have the effect of imposing a much lower discount for contingencies in wrongful death claims than in personal injury claims. Order

The appeal should be dismissed.

KIRBY J. In A A Tegel Pty Ltd v Madden , a decision which re-expressed one aspect of the law under consideration in this appeal in the light of changed social circumstances , McHugh JA invoked words of Frankfurter J in the Supreme Court of the United States. In Henslee v Union Planters Bank , Frankfurter J observed: "Wisdom too often never comes, and so one ought not to reject it merely because it comes late."

The present appeal is from a judgment of the Full Court of the Supreme Court of Western Australia . It concerns a claim brought by a widow under the Fatal Accidents Act 1959 (WA) ("the Act"). In her claim, the widow succeeded on the issue of liability . The damages recoverable by her for herself and her two children by the deceased had to be calculated. By the Act, such damages were required to be "proportioned to the injury resulting from the death to the parties respectively for whom and for whose benefit the action is brought" .

Upon that brief statutory expression, or words like it, hangs more than a century and a half of judicial exposition. It has gathered around the Act and similar statutes that, in most countries of the Commonwealth of Nations and in the United States , copied Lord Campbell's Act . Lord Campbell's Act had itself repaired a principle of the common law that had applied a rule "[w]ith singular harshness" that "the death of a human being could not be complained of as an injury" . That rule may originally have arisen from historical confusion . However, it required, and obtained, legislative correction. Such legislation applies in all Australian jurisdictions in similar terms .

The main issue in this appeal concerns a substantial question of legal authority, principle and policy . It is whether the computation of damages should take into account the best estimate that can be given of the prospect of re-partnering with economic benefits in the future. Regard is to be had to significant changes in social conditions in Australia since this subject was last examined by this Court and to closer analysis of the assumptions propounded to support a deduction in damages, in the absence of proof that the plaintiff has in fact remarried or formed some other equivalent domestic relationship involving actual financial support.

Three issues were argued in the appeal . However, the basic issue of principle is that just stated. The main question in the appeal is whether the Full Court (or the primary judge) erred in the application of that principle to the facts of this case. The facts and applicable legislation

The facts upon which the damages payable to Mrs Teresa De Sales ("the appellant") fell to be determined are stated in the reasons of other members of this Court . So is the course of the litigation in the Western Australian courts, the difference of opinion in the Full Court and the applicable provisions of the Act .

The primary judge concluded that a discount of 5 per cent for "the chance of obtaining support from remarriage" was appropriate . In the Full Court, Wallwork J would not have disturbed that approach . However, a majority of the Full Court (Miller J, with whom Parker J agreed) concluded that the discount allowed for "the prospects of remarriage" was inadequate and should be increased to 20 per cent . When combined with a further discount of 5 per cent for "general contingencies", the Full Court ordered that the appellant's damages be reassessed and reduced by 25 per cent .

In this Court, there is a divergence of view. Gaudron, Gummow and Hayne JJ conclude that no separate deduction should be made for the prospects of forming a new economically supportive relationship; that no such deduction should have been made in this case; and that, accordingly, the appeal should be allowed . Gleeson CJ would restore the deductions found by the primary judge . So would McHugh J and Callinan J, whilst expressing the view that the discount of 5 per cent was "unreasonably low" or "too little" .

In practical terms, these divergences suggest that I must choose between the two approaches. The first is one that maintains the legal principle that, in calculating damages under the Act, the judge or jury are bound (absent actual evidence of established or proposed re-partnering) to make an estimate of the future prospects that a plaintiff, who has lost a spouse or domestic partner, will enter an economically supportive substitute relationship in the future. The second consigns that approach to legal history and overrules the past authority that has accepted such a deduction. Legal context and past authority

Pecuniary recovery for wrongful death : The Act has been interpreted as limited to the recovery only of the economic or material advantages lost by the survivors . In this respect, Australian law has followed the course of judicial authority and commentary concerning its English progenitor , as well as the approach in other jurisdictions . As a matter of construction, it did not, in my view, have to be so. The statutory action is derivative from that of "the party injured" . The measure of the damages refers to the "injury resulting from the death" . For most mortals, the loss by death of the "relatives" for whom the Act provides would usually occasion shock, grief and distress. Some jurisdictions make a specific provision for solatium . The Law Reform Commission of Western Australia recommended such an addition to the Act . The recommendation was not implemented. Such "injuries" are therefore regarded as outside the statutory entitlement. It follows that, by settled law, not challenged in this appeal, recovery is restricted to the loss of dependency measured in financial terms.

Recently, the European Court of Human Rights has held that, by not providing an adequate remedy for non-pecuniary loss in respect of death, English law is defective when measured against the Convention for the Protection of Human Rights and Fundamental Freedoms . This ruling may lead to the amendment of English statutory law . However, it does not affect the meaning, and basic purpose, of the Australian legislation. The foregoing limitation illustrates the proposition that the remedies provided by the statute are fundamentally limited and imperfect. In Australia, the ultimate source of compensation in a case such as the present is, and is only, the statute.

Collateral benefits and burdens : Just as the character of the legislative provision for damages is settled, so too is the basic approach to the assessment of damages. The law requires that "any benefit accruing to a dependant by reason of the relevant death must be taken into account ... the balance of loss and gain to a dependant by the death must be ascertained, the position of each dependant being considered separately" . This follows from the requirement in the Act that the damages be proportionate to the injury to the dependants. The "cardinal point" of the legislation is its compensatory character . This being so, a court must endeavour to work out the balance of financial gains and losses as best it can .

In Bresatz v Przibilla , Windeyer J pointed out that "[a]ll 'contingencies' are not adverse: all 'vicissitudes' are not harmful." His Honour asked why the "buffets" should be counted but the "rewards of fortune" ignored. This kind of thinking is reflected in a large number of decisions by which courts have classified particular events as within, or outside, the benefits accruing to the relatives that must be brought to account in calculating the "injury resulting from the death".

Benefits from a public fund set up following the death ; help with the housework given by a third party ; and the payment of a pension under a statutory scheme have been held outside the benefits for which credit must be given. The law on the subject is full of anomalies and fine distinctions. Why, for example, should help given by a step-parent to a child who has lost a parent be deducted in the calculation of damages under the Act, but not similar support given by a grandparent or other volunteer ? Sometimes, the only apparent justification for the lines drawn by judicial decisions in this area has been that of a policy choice . The logic of several of the decisions is questionable, a fact recognised by this and other courts .

The discount for re-partnering prospects : For many years, one of the contingencies which the law in England and other Commonwealth countries required plaintiffs to bring to account, in reduction of the "injury resulting from the death", has been the prospects, as it was traditionally put, that the plaintiff will remarry.

There were at least two bases for the discount for possible remarriage. The first was the "revived capacity to marry", suggesting that the mere capability of remarriage, due to death, was of some inherent economic value, that the very freedom to do so was to be given a pecuniary figure . However, the prevailing basis for this approach focussed on the financial element of relationships; assumed dependency of a wife on her husband; and assumed that the death of a husband opened up the possibility that the widow would find another, substitute, husband able to support her in a way financially equivalent to the deceased. If such an alternative "bread-winner" were found quickly "a widow could reasonably recover but small damages when the death ... has been quickly followed by marriage to a man who, from his wealth, can provide much more than bread" .

This reasoning led to attempts to estimate the prospects that a deceased's spouse (usually a widow) would find good domestic and economic fortune. In the result, it was not uncommon to see widows described by judges as "well groomed, attractive and presentable ... [with] a personable and warm nature" and other such like epithets. On my reading of the cases, such an evaluation of the physical attractiveness is not normally made in the case of male claimants . If the judge considered that the widow was "elderly" or of "unattractive appearance or disposition, or suffers from some disability, or is encumbered with a large number of young children" such comments (except perhaps the last) might usually be left unsaid but with only a small discount for the prospects of remarriage being allowed. In the absence of an acknowledged or proposed remarriage , or established or admitted like relationship , trial judges would usually be left to their own estimations concerning the likelihood of such an event.

The intellectual justification for taking into account a widow's "revived capacity to marry", whilst ignoring her capacity to return to work, was said to lie in the fact that the latter was always available whereas the former could properly be classified as "resulting from" the death of the deceased . More accurately, in most cases it was not the prospect of remarriage, as such, which would diminish the pecuniary losses of the relatives concerned but the possibility that, with such remarriage, would come financial support reducing the damage otherwise resulting from the death . The approach inherited from the English cases has been followed in Australia for many years . The trial judge was expected to perform such factual inquiries . The legitimacy and feasibility of the inquiry was not challenged until now; it was simply assumed to be proper and necessary. In fact, this Court has previously stated that "it is profitless to debate how far the established rule is justified" , because of the historical acceptance of the deduction.

United States authorities : In the United States, legislatures copied Lord Campbell's Act in much the same way as they did in Australia. However, on this particular issue of suggested offsetting benefits, the courts of that country, virtually without exception, forbade account being taken of the fact of remarriage of a deceased's spouse prior to the trial . Still less were estimates of the prospects of remarriage at some time in the unknown future permitted to reduce the damages . From an early time, a strong line of authority accepted that the entitlements of relatives under the Act were not lost, divested or reduced by the consideration of remarriage, actual or estimated .

At various times over more than a century during which this approach has been taken, different explanations have been offered by the courts of the United States for adhering to it. Such explanations have included: (1) As the loss under the statute for which the relatives sue is fixed at the time of death, events happening thereafter are immaterial ; (2) For policy reasons the tortfeasor ought not to be allowed to rely upon fortuitous events mitigating the relatives' economic losses, to which the tortfeasor has made no contribution ; (3) Excluding such considerations helps avoid resort by plaintiffs to untruths about their personal relationships, which they should not be obliged to conceal, modify or perjure themselves about ; (4) The exclusion of such considerations, in an action already limited to the recovery of economic loss, is justifiable as maximising the recovery of a widow to whom the law was generally sympathetic ; and (5) In any case, the calculation of the prospects of securing an economically advantageous remarriage (or equivalent relationship) was too problematic in the individual case, taking the courts into the realm of hopeless speculation rather than an accurate judicial calculation of the economic damage resulting from the death as contemplated by the statute .

Reason (1) can be disregarded. It is clear law that, in Australia, proof of remarriage with beneficial economic consequences (or the establishment of an equivalent relationship) prior to trial may be taken into account under the Act on the footing that the law prefers calculation to speculation wherever the former is available . As to (2), (3) and (4), these seem unconvincing reasons in a legal system committed to even-handedness between plaintiffs and defendants. However, reason (5) has more substance. It formed the basis of the appellant's submissions in this Court.

The basic question in this appeal is therefore stated as follows: Is it too late for this Court, in judicial proceedings, to re-express the common law or, more accurately, the application of the provisions of the Act, in circumstances where the "damage" is said to be likely to be reduced by the possible achievement of a future stable domestic relationship with long-term economic benefits that defray the economic "damage" resulting from the death? Even if, belatedly, this Court were to agree with the United States courts that such reductions involved hopeless speculation, should the law remain as it is, leaving it to the legislatures to change the established principle if they chose to do so? Arguments for adhering to re-partnering deductions

Before explaining why I believe that re-expression of the law is both permissible and necessary, I must acknowledge that there are a number of arguments that I have taken into account against that course.

It is insufficient to abandon the past law on this topic simply because it requires a measure of speculation . It has been said many times that the one certainty about the calculation of damages in personal injury claims, and those brought under the Act, is that the estimates made, and the hypotheses upon which they have been calculated, will in time be shown to have been erroneous . The fact that the tools available for decision-making are imperfect is not a reason for abandoning the attempt. That would be to confuse "the measuring rod for the thing to be measured" .

The assessment of the "damage" and the "proportioning" of it in terms of the "injury resulting from the death" inescapably involves many hypotheses and estimates. This is inherent in the system of once-only verdicts . Insurance assessors, economists , actuaries and other citizens must make judgments about future imponderables all the time. There are many "conjectural questions" in the computation of virtually every damages verdict .

Whilst conceding that calculation of the prospects of re-partnering may be unscientific and that courts should not deceive themselves into false beliefs about the possibilities of precision , the general guiding rule for the calculation of damages remains that courts should proceed "as accurately as possible ... [if necessary] with the assistance of actuarial material" . The general endeavour should be to reduce the guesswork in favour of rational estimation, not the needless abandonment of that effort simply because of the difficulties of executing it . If the purpose of the Act is to compensate relatives for the financial injury resulting from a death, and if the common experience of humanity is that such financial loss will sometimes be reduced because a survivor is able to secure economic support from another domestic partner, the lack of certainty that this will occur in a particular case, or to what extent events will happen, is not necessarily a reason for giving up the effort to make the best estimate possible on the available data so long as the resulting calculation is more than unreliable guesswork or fanciful speculation .

The proposals of law reform bodies to abolish this deduction are not themselves reasons why a court should necessarily do so. Such bodies speak to the legislature which establishes them and which has the power and legitimacy to adopt or reject their proposals. Only one Australian legislature, in the Northern Territory, has enacted a law for the abolition of the "remarriage" deduction, and then in a limited way . However, there are a number of overseas legislatures that have taken this course . In any event, the reform proposals have been criticised . The respondent urged that action upon them, if any, should be left to the several Parliaments, which are accountable, as this Court is not, to the electors. The failure of most such Parliaments to adopt the reform proposals presented to them was itself said to be significant .

This Court needs to observe its limited function in any re-expression of the law. It should not lightly undermine settled rules of law that have long been followed. It should avoid any approach to the computation of damages in personal injury claims unduly sympathetic to plaintiffs , of the kind that appear to be reflected in at least some of the arguments adopted by courts in the United States for adhering to the rule that forbids account being taken of the possibility of domestic re-partnering. Judicial complaints about the deductions

Judicial complaints : Accepting the force of the foregoing reasons for adhering to the previous approach to this subject, two aspects of the judicial treatment of the discount for re-partnering must be specially noticed.

First, for as long as judges have been required in England, Australia and other countries, to estimate a discount for this consideration, many have complained about the obligation to do so. The complaint represents a particular species of the not uncommon judicial reprise about the quest for the certainty of arithmetical calculations, based upon the false assumption that it is possible to calculate damages in personal injury claims with a high degree of accuracy and precision .

In the particular field of deduction "for the prospects of remarriage" judges have declared that they were confronted with many "matters of speculation and doubt" . They have objected to being required to estimate such matters "incapable of evaluation", which at best could be "but roughly calculated" . They have declared that such a duty subjected them, or a jury, to a "guessing game" . They have acknowledged that the discount was "always difficult to assess" . It carried with it serious risks of injustice to a particular plaintiff who, in the events that actually transpired, did not live up to the judge's optimistic prognostications. In Nunn v Cocksedge Ltd , Denning LJ observed, in words that are particularly apt for this case: "[I]t seems to me that the judge must have reduced his figure far too much on that account. After all, she has not remarried. She is not even engaged to be married. Two years have passed since the accident, and it is entirely an element of chance whether she will remarry or not. How sad it would be if, as the years passed by, she did not remarry, and yet her compensation had been cut down by the court so drastically on the footing that she would!"

Secondly, in more recent years, judges have begun to criticise even more forcefully what many have seen as a "distasteful", "invidious" and unseemly inquiry and assessment on this score . In his reasons, Callinan J expresses the view that the task of assessment of the prospect of re-partnering is no more "distasteful" than many other duties of assessing damages, eg for injuries to sexual functions . I agree that such subjects, which can be very important to particular plaintiffs, should be examined openly and honestly in the courts . However, it is one thing to scrutinise evidence about such subjects where a current claim is made about them. It is quite another to subject an individual to distress, humiliation, investigation and "dirt digging" where the person involved may be vulnerable and quite uncertain about present and future personal relationships .

In one recent case, the judge described the defendant's cross-examination on this issue as having "caused [the plaintiff] great distress ... [it] underlines the humiliating nature of the task of assessing the proper discount for the prospects of remarriage ... [The plaintiff] was heavily drugged with valium to help her cope on that day, and her remarks only give insight into the fact that she was distressed and as one might expect, less coherent than usual" . The stated experience of trial judges in this country, in England and elsewhere suggests that this Court should not lightly dismiss such objections to which even hardened judges have made reference.

Judicial inconsistency : In addition to this consideration, regard must be had to the very great disparities that appear in the calculations to which individual judges come when estimating the discount that should be allowed for the prospect of re-partnering. In some cases, judges have suggested that too great a weight must not be given to this factor and that the normal allowance should be moderated . In one recent case of this type a discount of only 2 per cent was provided . In another, the discount for all contingencies including the "prospects of remarriage" was 10 per cent . In this Court, Menzies J in Jones v Schiffmann favoured a discount that was equivalent to 17 per cent. Discounts of 20 per cent have not been uncommon in State courts . In a case where the judge considered the plaintiff "personable and warm", the discount allowed was 50 per cent . In another case, where the trial judge had allowed only 20 per cent for all contingencies, the Full Court of the Supreme Court of Western Australia increased the discount to 60 per cent . In Willis v The Commonwealth , where the widow's remarriage within a short time of the death of her husband was actually proved, the discount was equivalent to 100 per cent. In contrast, in a recent case involving a widower, the judge came to the view that there were no prospects of remarriage and therefore made no deduction .

It might be argued that these variations merely demonstrate the infinite variety of circumstances proved in, or inferred from, the evidence of a particular case. However, another explanation may be that the estimation depends upon imponderable factors, that it relies too much on considerations of the personalities and attitudes of the judges or juries, typically after a very short encounter with the plaintiff, when they engage in the re-partnering "guessing game" . The changed circumstances

Changed social circumstances : Since the requirement to provide the discount that is in contest in this appeal was first accepted by the law, the social assumptions upon which it depended have changed radically.

Amongst the changes are the following: (1) The decline in the number of domestic partnerships formalised in marriage and the significant increase in de facto relationships ; (2) The growth in the incidence of divorce, making past assumptions about the permanency of life-long economic support by a domestic partner much more doubtful ; (3) The growing incidence of female employment and of the economic and social independence of women ; (4) The availability of social security payments that render individuals less financially dependent than they previously were upon domestic partnerships with another person ; (5) The growing number of short-term domestic relationships, the decline in monogamy and the incidence of same-sex relationships; (6) The reduced social stigma affecting single women who elect to live without a domestic partner; (7) The greater mobility between economic classes both of men and women, which challenges the assumption of re-partnering with a person of similar economic capacity to that of a previous partner; (8) The increase in the number of judges who may be less likely to refer to the physical attractions, warmth of personality and remarriage prospects of a widow based upon stereotyped assumptions about the considerations that contribute to the initiation and continuance of domestic partnerships ; and (9) The longer life expectancy of the population, the increasing incidence amongst domestic partners, as other persons, of the diseases of the elderly and the economic burden that old age can place upon families and domestic partners.

These considerations make the factual foundations for the assumptions and calculations about re-partnering to economic advantage, that were thought possible fifty, forty and even thirty years ago, much more problematic in contemporary Australia. The remarriage tables relied on thirty years ago are now unreliable. In any case, they neither show the intervals between marriages nor the nett economic support of a later marriage, nor do they show non-married domestic partnerships or other relationships that may involve physical and emotional support without any economic underpinnings or consequences.

Where the assumptions upon which previous statements of the common law or of equity or the previous construction of legislation have changed so radically, a time arrives when courts, and particularly this Court, must alter their approach in order to escape the justifiable criticism that they are perpetuating expressions of the law that are anachronistic or impermissibly discriminatory .

Our law has moved a long way since Blackstone asserted: "the very being or legal existence of the woman is suspended during the marriage, or at least is incorporated and consolidated into that of the husband: under whose wing, protection, and cover , she performs every thing" . Yet reading the cases on the so-called "remarriage discount", one cannot escape the conclusion that they reveal a "distinctly male perspective" . When this conclusion is reached, it is essential that this Court should re-examine the assumptions that underlie previous expositions of the law and, if so warranted, re-express that law in a way that is more harmonious with contemporary legal principle and social reality.

Excessive speculation : When to the foregoing considerations is added the number and variety of variables that, in a particular case, will affect the prospects of achieving economically beneficial re-partnering, a powerful reason is provided to resolve the re-examination of the discount by abolishing it.

The variables are affected by estimations about the deceased, the deceased's spouse, partner or family member, and any future spouse(s) or partner(s) with whom an economically relevant relationship is formed following the death. Concentrating only on the last-mentioned person(s), the variables require estimates to be made about the interval before which a new relationship begins; the economic features of that relationship; its duration; the possibility of its breaking down; the risks of supervening unemployment, illness or death of the new partner and replacement with a further relationship in which all of the same variables arise again. Such estimates must then be adjusted not simply by reference to any table of averages but having regard to the peculiarities of the personalities of those involved. Life expectancy, viewed alone, may be fairly calculated by reference to statistical tables. However, the endurance of personal relationships is prone to the unpredictable vicissitudes of human personality, desire and fortune.

Even more fundamental is the error of assuming that the result of the formation of a new relationship will be economically positive. On the contrary, it may be detrimental to the finances of the deceased's spouse or partner. The new partner may be without any assets, unemployed and dependent on the claimant. Without evidence as to the specific relationship, there is no way of knowing whether the relationship will be of benefit to the claimant or not.

I acknowledge the force of the argument that this Court should leave change in this area of the law to Parliament. On occasion, especially where the law in question concerns criminal liability or has large economic implications, I have favoured that view . However, where judge-made law is shown to be anomalous or overtaken by changed social conditions, this Court has not hesitated to re-express the law. It has done so even where some legislative changes have already been made upon the assumption of the correctness of previous understandings of the law .

The issue having been squarely presented for decision and argued in this appeal and the present approach having been shown to be unjust, unpredictable, anomalous and discriminatory, the time has come for re-expression of the law on the discount for domestic re-partnering. I therefore agree with the joint reasons that, in a wrongful death case, ordinarily, no deduction should be made on account that a surviving spouse or domestic partner will remarry or form a new domestic relationship of economic significance.

Even where there is proof of such a relationship, that fact must be considered in light of all the circumstances of the case. As has been noted both in Australia and overseas, "[i]t does not necessarily follow that if a widow re-marries ... her right to financial support from those who killed her husband necessarily comes to an end" . Notwithstanding the fact that the level of economic benefit is found to be greater than that provided by the previous relationship, other contingencies have to be taken into account just as they are in determining any amount of damages relating to the future . Deductions for general contingencies

One significant problem arises tangentially in this appeal. It was disclosed by the course of the argument. It concerns the disparity that has emerged in different Australian jurisdictions in relation to the deduction for general contingencies made to an award of damages in tort . In New South Wales, a standard deduction for contingencies of 15 per cent was noted, without adverse comment, by four Justices of this Court in Wynn v NSW Insurance Ministerial Corporation . A criticism of the practice of awarding a standard deduction regardless of individual circumstances, as expressed by me in the Court of Appeal of New South Wales, was mentioned briefly in this Court but without elaboration . In Western Australia, relevant to this case, the standard deduction for general contingencies is between 2 and 6 per cent .

This disparity between the approaches taken to "general contingencies" in different States may one day need to be considered by this Court. On the face of things, there would appear to be no justification of legal principle for such a wide divergence in the rule applied by courts in different parts of Australia. However, that issue is separate from, and different to, the one that was argued in this appeal. To resolve such inter-State disparities, this Court would need to consider comparative materials, a wide range of relevant contingencies and the approach (if any) that should be substituted for them all. The issue presents a question as to whether a semi-arbitrary rule of thumb can be justified on grounds of realism, convenience and economy or whether it is necessary to evaluate the evidence of each particular case for "rational principles upon which damages ... are to be assessed" . Because this Court did not hear full argument on that question in this appeal, I would refrain from saying more about it. The larger, and different, question of the discount for general contingencies of life must be left to the future.

No discount for general contingencies was allowed in this case by the primary judge. On appeal such a discount was fixed at 5 per cent. This Court was told that this fell within the standard range of such discounts made by courts in Western Australia. It seems rather low, compared with the standard discount in other States, such as New South Wales. However, I am willing to agree with the joint reasons and uphold the Full Court's conclusion on this point, pending any comprehensive review of that question. Conclusions and orders

In the result, what was formerly the discount for "the prospects of remarriage" does not apply, in the absence of evidence of actual remarriage with beneficial economic consequences, or the actual or proposed establishment of a like relationship. The accurate assessment of economic benefits and losses from a hypothetical future relationship has been shown to be impossible and undesirable. What has been known as the discount for the prospects of remarriage is therefore no longer part of the law.

The only remaining question is whether the possibility of hypothetical re-partnering should be treated as subsumed within the "standard" discount, where it exists, for what are called the general contingencies or the "vicissitudes" of life .

There is a practical risk in taking this course. Unless care were observed there could be a reduction in the transparency of the process of calculation. This appears to have occurred in Queensland, where a practice seems to have emerged of giving a single percentage figure as a total discount, combining therein a discount for general contingencies and one for the prospects of re-partnering . Such an approach could produce an effectively unreviewable deduction. It would represent one step forward and two steps back. It is not a course that I could favour.

On the other hand, in many parts of Australia, there is a judicial practice of allowing a standard discount for all of the unquantifiable contingencies that may occur in the future. Relevant to this case, Western Australia is one of them. Such discounts themselves present problems of principle . However, I acknowledge that some such discount is normally appropriate to avoid overcompensation resulting from the mechanical application of multiplication tables to a present loss. To the extent that the economic advantages or disadvantages of hypothetical re-partnering are relevant to the calculation of the "injury resulting from death" they should now be taken as included in the "standard" adjustment for imponderable future considerations. But no attempt should be made by the judge to evaluate more accurately the possibility or probability of such an outcome. Courts should not pretend to such an unattainable standard of predictive certainty. Re-partnering is merely another of the many possible vicissitudes of life, namely that the claimant may enter an economically beneficial or detrimental relationship after the trial. It is therefore to be given no more weight than any of the other vicissitudes that go to make up the general discount. The "standard" adjustment should not be increased to re-introduce the "remarriage" discount by the back door.

These reasons do not address whether, or the way in which, this particular contingency could be subsumed into the discount for general contingencies in those Australian jurisdictions that do not have a "standard" discount. That issue must await further consideration, together with the issue of the disparity in the approaches of each State in relation to deductions for vicissitudes more generally .

I concur in the orders proposed in the joint reasons.

CALLINAN J. The two questions which this appeal raises are whether any deduction should be made from a widow's damages for loss of support by reason of her prospects of remarriage, and if it should, what is its appropriate measure. Facts

I use the expression "loss of support" in these reasons for convenience, and because it has been used historically to embrace not only support provided by way of maintenance, but also other benefits having a monetary value. Similarly, I refer to prospects of remarriage as an abbreviation for prospects of remarriage or another relationship from which benefits having a monetary value may result.

The appellant was 27 years old in 1990 when her husband died in an accident in circumstances which entitled her to sue on her own behalf and on behalf of their children for damages for the loss of support which would, but for her husband's death, have been provided to her and her children who were then aged about 3 years and 4 months respectively. The appellant's husband was 31 years old when he died, and was a graduate and an accountant by profession. At the time of his death the deceased was employed as a financial controller for a group of companies.

The appellant had worked and studied for periods during the marriage. She was not working at the time of her husband's death. Since then, she has worked full time in various occupations. She has not remarried and she has no plans to remarry, although she has not resolved against it. The appellant has had "a purely physical relationship" lasting about three and a half years with another man since she lost her husband. She gave these reasons why that relationship did not lead to marriage: she did not love him; the man's children did not want to form a close relationship with her; and she "had learned to become very independent". The appellant intends to move to another State. Her aim is to educate her children until they graduate from university and then to travel and study in Europe. Her daughter has expressed a wish to be a teacher. Previous proceedings

The appellant brought her claim in the District Court of Western Australia on her own behalf and on behalf of the children of the marriage. HH Jackson DCJ who tried the issue of damages assessed them (before apportionment for contributory negligence) as follows: "As I have said, in my view a modest reduction should be made for the chance of obtaining support from remarriage, and I deduct five per cent. The deduction should only be made from the share of award that is apportioned to the plaintiff. It follows that the award should be formulated prior to apportionment as follows: Tara - for loss of pecuniary benefits $60,000 for loss of parental guidance $10,000 $70,000 I allow interest on that sum as follows: $70,000 x 4/100 x 9 years = $25,200 $25,200 $95,200 Jordan - for loss of pecuniary benefits $64,700 for loss of parental guidance $10,000 $74,700 I allow interest on that sum as follows: $74,700 x 4/100 x 9 years = $26,892 $26,892 $101,592 Mrs De Sales - For loss of pecuniary benefits (from the sum of $665,850) must be deducted awards of $95,200 and $101,592 to Tara and Jordan, leaving a balance of $469,058. From that amount must be deducted a contingency of five per cent for the prospect of remarriage, leaving $445,605. I add to that $8,650 for funeral expenses, a total of $454,255. I allow interest on that sum as follows: $454,255 x 4/100 x 9 years = $163,531, a total award to Mrs De Sales of $617,787."

In making his assessment his Honour was obliged, as all assessors of damages in personal injuries cases and this sort of case are, to make predictions with respect to the future. They included that the appellant and her husband would have continued to pool their incomes and expenditures; that in the future the deceased would have made available 75 per cent of his nett earnings to the appellant and the children; that the children would have had a call upon the deceased which he would have met, for virtually full support until they attained 22 years; and that each child would have received from the deceased "parental support, guidance and training" quantifiable in money of $10,000. Involved in these predictions, or, as they become in a damages case, findings, are several assumptions: that the appellant and her husband would have remained married until the deceased reached 65 years; that the degree of dependence would have remained unchanged throughout this period; that the deceased would have had a career that would have given him the capacity to provide the support that the judgment contemplated; and that the children would not have become self-supporting or otherwise independent before attaining 22 years.

As a basis for some of his predictions the primary judge relied upon tables and statistics: for example, results of surveys of salaries compiled by the Australian Institute of Management; and tables as to financial provisions within households and life expectancies prepared by an actuary from the statistics gathered by the Australian Bureau of Statistics. These statistics and tables were either proved in evidence or were uncontroversial. Neither party suggested that they were not relevant despite that in practice few individuals or particular situations will conform to the average. Neither party however set out to prove, or sought to rely upon any past or current statistics with respect to marriage breakdowns, remarriages, or the formation of de facto relationships. Other bases for his Honour's predictions were the history of the appellant's and the deceased's marriage and working lives.

The appellant appealed, and the respondent cross-appealed against the judgment of the District Court to the Full Court of the Supreme Court of Western Australia . With respect to the questions with which this Court is concerned the Full Court was divided. Wallwork J said this : "In this case, in my opinion, the learned judge was entitled to take 5 per cent from the damages for the contingency of re-marriage. It was a minimal sum and I personally would have set it higher. That is because the appellant was relatively young and very capable with two children who would not take long to reach adulthood."

Miller J, with whom Parker J agreed, was of the opinion that a discount of 5 per cent for "prospects of remarriage" was too little. His Honour's view was that, having regard to the "appellant's age and credentials" a deduction of 20 per cent was appropriate . To that percentage his Honour added 5 per cent for "general contingencies" and reduced the appellant's damages as reassessed by the other members of the Full Court by 25 per cent. The appeal to this Court

The appellant makes three basic submissions: that discounting for prospects of remarriage should be abolished; alternatively that this Court should provide guidelines which have regard to "modern day realities", as to appropriate percentages of discount; alternatively that the Full Court erred in interfering with the trial judge's assessment of 5 per cent as a contingency for remarriage.

The appellant contends that there are several reasons why "discounting for prospects of remarriage" should be abolished. There is little certainty in the approaches of judges. The appellant provided examples . The assessment of an appropriate discount is a very imprecise exercise. The task is distasteful to judges and demeaning to plaintiffs. In other jurisdictions discounting for a prospective remarriage has been abolished by statute , and, by analogy, this Court should do so . In the light of contemporary realities such discount (if any) as might be made should be upon the basis of likely financial benefit as a result of remarriage or otherwise, rather than on the basis of possible remarriage simpliciter. ( Jones v Schiffmann was said to be an authority of this Court for the erroneous course to the contrary which has been followed). Even if the touchstone is not remarriage per se, but the future formation of an apparently permanent relationship with possible financial consequences, because so many uncertainties arise as to the permanence of the relationship, no deduction for it is appropriate. Implicit in all, and explicit in some of the submissions was an assertion that the times have changed, and with them the communal approach to marriage, relationships and child rearing, which render past decisions and all assumptions as to these matters obsolete: and, because they are obsolete, entirely irrelevant. In their place should be substituted by this Court a judicial policy of denying any discount for remarriage or a future financially beneficial relationship.

The appellant's submissions must be examined in the light of the enactment which governs the appellant's action, the Fatal Accidents Act 1959 (WA) ("the Act "). The genesis of the Act is the Fatal Accidents Act 1846 (UK) ("Lord Campbell's Act"). Sections 4 and 6 of the Act provide that any action brought under the Act shall be for the benefit of the wife, husband, parent, and child of the deceased. Section 6(2) speaks of "damages ... proportioned to the injury ... to the parties respectively ... for whose benefit the action is brought". That "injury" means "financial injury" follows not only from the earlier reference in the section to "damages" but also, and particularly from the references in s 5 to the several, possibly beneficial, financial consequences of death for a dependent survivor which are to be disregarded for the purposes of assessing the damages. Injury has always, and rightly, been so understood in the cases, as financial injury, measured by the value of the support that would have been provided by the deceased to his or her dependents. In the words of Pollock CB in Franklin v The South Eastern Railway Company "[the damages] should be calculated in reference to a reasonable expectation of pecuniary benefit, as of right or otherwise, from the continuance of the life" .

The submission that discounting for the prospects of remarriage or some other relationship should be judicially abolished should be rejected. None of the reasons in support of it are persuasive so long as it is clearly understood that the discount is not to be made for the possible remarriage itself, but for the financial benefits that could reasonably be expected to flow from it.

Assessments of damages for both personal injury and loss of support are all necessarily imprecise because they have to be predictive about notoriously unpredictable matters, human affairs. In the interests of finality, not only in this field, but also in many others, the law requires that damages be assessed and paid once and for all, as a lump sum, even though the future, if the assessment were to await it, might falsify the assumptions underpinning it. Imponderables abound. They are certainly not confined to prospects of remarriage. As I have pointed out, the primary judge here had to make many assumptions about what the future might have held, and what it would in fact hold in the dramatically changed circumstances of the deceased's death: for example, whether even this apparently sound marriage would have endured; that the deceased's capacity to provide support had he lived would have increased rather than have been reduced by changing economic circumstances, shrinking opportunities of employment or other cause; whether the appellant would have ceased, because of enhanced opportunities of employment, of which she would have availed herself, to be dependent upon the deceased; and whether the deceased might have been disabled or struck down by some other non-compensable event, rather than the one which gave rise to this case.

Damages in cases of this kind are to be assessed by reference to the actual circumstances of the family, as they were, and can best be ascertained, before the death of the provider, and of the survivors in the future, as best they can be projected. They are not to be assessed arbitrarily. They are to be assessed in a way that does justice to both the plaintiff and the defendant. It would be unfair to a defendant, and would make a mockery of the law if a judge were required to blind himself or herself to a beneficial remarriage or, for example, the acceleration of a large inheritance from the deceased if that had occurred, or were to occur by the time of the trial. Kober convincingly refutes the arguments in support of the decision of the United States District Court for the District of Massachusetts, Wyatt v Bonnell Co in which the Court refused to receive evidence that the widow had remarried. The basis for justifying the exclusion most heavily relied on (and repeated here) is that the court is required "to embark upon a realm of speculation and be led into a sea of impossible calculations" . An assessment which disregarded the realities of the vicissitudes of life would do justice to neither party. As Windeyer J said in Parker v The Commonwealth , not all of the vicissitudes of life will be adverse to the plaintiff. It would, however, be to do a real injustice to a defendant to assess a widow's or widower's damage as if she or he would not remarry, even though it was the fact that a prospective spouse had the capacity to, and would be likely to support the surviving former dependent on a scale far beyond that provided before the death of the deceased. As best it can be, the likelihood of the occurrence of the various contingencies of life, will be assessed by reference to the evidence of the circumstances. To some extent human experience and judgment will need to be called in aid in the making of the assessment. So too will be the impression formed by the trial judge of the survivor. The fact that cases may be pointed to which indicate that on different occasions judges have adopted varying, indeed even widely varying percentages of discount does no more than demonstrate that cases and situations can depart significantly from any supposed norm. There is nothing unusual about the process involved. Judges (and juries) have undertaken it for many years and do so every day. Because a process is neither scientific nor productive of precise outcomes, is not a reason not to adopt it .

One writer in the United States summarized the position in a manner with which I agree : "If a jury is competent to determine the economic loss to the surviving spouse on the basis of such conjectural matters as life expectancy, probability of continued economic productiveness of the deceased spouse, and probability of continued matrimonial harmony between the deceased and the surviving spouse, they should also be competent to temper this figure with the probability of economic support from a new spouse that the surviving spouse would not have if the decedent had not been killed."

Much of the primary judge's reasoning in this case depended, and correctly so, upon the circumstances of the deceased's family before his death: the family's history was thought by the primary judge to provide a reasonable guide for its future. Part of that history was that the appellant had made, and enjoyed a happy, fruitful and apparently stable marriage, and, as she had not resolved against remarriage in the future, the possibility that she might, as an intelligent, companionable, selective person who had enjoyed a happy marriage, do so again, simply could not, in the trial judge's opinion, be ignored. To ignore that possibility would be to disregard human nature, a natural desire for companionship, that people do remarry and form enduring relationships (as the appellant once had), and all the elements of attractiveness, of mind, spirit, personality, intelligence, and appearance of which this appellant was apparently possessed: in short to ignore reality. There is nothing unique in the evaluation by a court of a party's assertion that he or she does, or does not intend to take, or would or would not have taken, a particular course. Many medical negligence cases call for precisely such an evaluation .

The exercise of looking closely to aspects of the personal life of a dependent should not be regarded as distasteful. The courts are concerned daily with intimate matters and relationships. The Family Court in particular, even today, is sometimes required to delve into these, in cases in which, for example, the interests of children are under consideration. Disabled plaintiffs whose sexual activities have been impaired give evidence in chief and are liable to be cross-examined about the extent of the impairment, and its effect upon their personal lives. The disfigurement of injured persons has frequently to be described in judgments and evaluated in monetary terms. Judges have to make judgments based upon the evidence. There are far more distasteful tasks than of assessing prospects of remarriage or of another relationship that have to be undertaken by judges: for example, the task of dealing with cases in the criminal jurisdiction concerning the most debased of human conduct; on occasions, of sentencing persons to terms of imprisonment in circumstances in which the sentence will cause great distress and deprivation to the family of the offender; of applying a law of which the judge strongly disapproves; and of having to find that a party has lied. That a judge might find a task distasteful is not a reason for the judge not to do it.

A finding by a judge as to a person's qualities and disposition to remarry, is not calculated to be demeaning of that person. And a judgment, sensitively expressed that remarriage is unlikely, or is a remote or non-existent possibility, but which accordingly results in a higher award of damages should not be unduly demeaning to the person with whom it is concerned. Many judgments may be demeaning to a party. A judgment holding that a defendant has been negligent, or indeed that a plaintiff has been guilty of contributory negligence may be demeaning to the person about whom the finding has been made. The finding which resulted in a higher judgment to the plaintiff in Gillies v Hunter Douglas Pty Ltd , that the second husband was an inferior provider to the plaintiff's first husband, may not have pleased the latter, but was one that had to be made and was for the plaintiff's benefit. Furthermore, the plaintiff there apparently had no qualms about advancing the argument that led to the judgment even though it reflected adversely upon her current husband. That a party may feel demeaned by a judgment of a court is not a reason for a court not to do its duty. If it were, courts would be seriously inhibited in performing their task of quelling controversies.

I would reject the argument that because some Law Reform Commissions have recommended, and in some jurisdictions it has been enacted, that there should be no reduction for prospects of remarriage, provides reason for this Court to adopt a like rule. Abstention from legislating provides a reason to the contrary. That there has been legislation elsewhere, shows, if anything, that legislation by a legislature, rather than its de facto equivalent by this Court, is the appropriate means of change if change be required, particularly in respect of what has become a longstanding conventional practice in cases of this kind. Moreover, for the reasons that I have already stated, I think that pronouncements in judgments and texts condemning the assessment of prospects of remarriage as distasteful and demeaning seriously overstate the position.

Something needs to be said about the appellant's submissions as to changing times and attitudes. On the opening page of Mansfield Park Jane Austen wrote in 1814: "[t]here certainly are not so many men of large fortune in the world, as there are pretty women to deserve them." It is easy to see that the author's pronouncement is discordant with contemporary Australian society and values. It is less easy however for a judge to say what those values are, and what part they should play in the application of a form of an enactment which has fairly consistently been applied throughout the country for many decades now. The appellant's submissions about change might have been helpful and more persuasive had they been supported by comparative statistical tables properly proved in evidence demonstrating how much, and by what margins changes have occurred. Their absence is all the more remarkable by reason of the proof and reception of other statistics at the trial. There is no reason why, as happened regularly after the publication of remarriage tables in 1971 , a judge at trial might not use relevant, reasonably contemporary, duly proved or admitted and sufficiently refined, properly compiled tables as at least a starting point for a consideration of the prospects of remarriage, or another permanent or enduring relationship. Courts should be wary of broad submissions about change. It is a primary responsibility of legislatures to identify change and react by legislating, when appropriate, rather than courts. In this case the appellant argued that fewer marriages were contracted today than in 1846 when Lord Campbell's Act was enacted. That may be so. But if it is so, the changes should have been proved by the tendering at the trial of comparable statistics (to the extent that reliable statistics were available in the 19th century). Any discussion would also have had to take account of the customs of the period, and the difficulties and expense of obtaining a divorce during it, as well as other, relevant, historical features. It may also be open to question how reliably judges and lawyers generally can make informed assumptions about social conditions, changes and practices and whether the changes are real or only apparent, and how widespread their effect is . One change that may have occurred, I cannot say whether it has or not, is that many women, of which this appellant may be one, transform their lives as their children grow older, by studying and working and ceasing to be dependent at all upon their husbands: indeed they sometimes become the, or the principal provider. A contemporary social condition not to be overlooked, if the court were entitled to consider it, might be the difficulty currently being encountered by people and organizations in obtaining affordable insurance, a difficulty to which very large assessments of damages by courts may have contributed: a reason itself for moderation, by having proper regard to relevant discounting factors.

I would emphasize however that statistics can only be a starting point. Statistics should only be considered in the light of the evidence in the case as to the more important, indeed probably in most cases, the critically important factors, of disposition and inclination of the surviving spouse to remarry or form a relationship, opportunity to do so, and its likely financial advantages, if any, to that survivor.

I would therefore accept only one of the appellant's contentions in relation to the first two submissions. It is important that the emphasis not be misplaced. If a court were to reduce the damages simply because of some prospects of remarriage or of another relationship, the court would be in error. What the court should have regard to is the prospect that the surviving dependant, may in the future remarry or enter into a relationship which may improve, replace, or go some way for some period, towards replacing, the financial benefits lost by reason of the death of the provider of them. It may also be that a prospect of lawful marriage might be more beneficial than of another relationship, because more substantial legal rights may sometimes arise out of the former . This may not be able to be done precisely, but it should not be done arbitrarily, and it cannot be done by using some guideline as a touchstone. It has to be done however by reference to the actual circumstances, some only of which may be the personal attributes of the survivor, male or female, and especially disposition and opportunities, as assessed by the judge or jury whose task it is to assess the damages in a particular case.

In summary, discounting of the kind which was done here should not be abolished. Guidelines are impractical and neither can nor should be introduced. The emphasis should be upon what a new relationship, if any, is likely to produce, in financial benefit. The survivor's disposition to remarry, and opportunity to do so are very important factors.

The discount should, if it is to be made, be made separately and specifically, and not as part of a general discount for contingencies. I say this for two reasons. If it is identified it will be examinable on appeal. Courts should strive to produce examinable, transparent judgments. Indeed there may be a strong case for the expression of a percentage for each factor that a judge regards as a discounting one whenever that is possible. The second reason is that, although each case is unique, as happens with assessments of damages for pain and suffering, a pattern tends to emerge from a number of judgments to provide a basis for future judicial assessments, and, in turn, settlements .

Should the Full Court have interfered here? I would myself have been inclined to think a discount of 5 per cent too little. It hardly seems likely that in the decades after the appellant's husband's death she would not enter into a relationship or relationships that would give her only 5 per cent of what the deceased would have given her had he lived, that is of approximately $30,000 in present value. But the assessment was one peculiarly for the trial judge. The Full Court neither saw nor heard her give evidence. A perusal of the record provides an incomplete basis for assessing whether the appellant had a disposition towards forming a financially supportive relationship, and was the kind of person who would have an opportunity of doing so. There is no suggestion that the primary judge failed to take into account any relevant factors or took into account irrelevant ones. Accordingly the Full Court should not have substituted the discount under this head that it did. However the matter should be remitted to the Full Court for further consideration in the light of the reasons of this Court for judgment as it is not clear how, if at all, the restoration of a discount of 5 per cent only, might affect the allowance for contingencies and the assessment generally.

I would allow the appeal with costs and order that the matter be remitted to the Full Court for further consideration.