THE SUPREME COURT OF APPEAL OF SOUTH AFRICA
JUDGMENT
REPORTABLE
Case No: 453/2011
In the matter between:
LIZE-MARI MACDONALD FIRST APPELLANT
PETRUS MACDONALD SECOND APPELLANT
SUMé MACDONALD THIRD APPELLANT
v
THE ROAD ACCIDENT FUND RESPONDENT
Neutral citation: MacDonald v Road Accident Fund (453/2011) [2012] ZASCA 69 (24 May 2012).
Coram: Navsa, Brand et Mhlantla JJA
Heard: 11 May 2012
Delivered: 24 May 2012
Summary: Dependants’ claim for loss of support – application of Assessment of Damages Act 9 of 1969 – calculations based on assumptions and contingencies not appropriate where actual figures available.
________________________________________________________________
ORDER
________________________________________________________________
On appeal from: Western Cape High Court, Cape Town (Bozalek J sitting as court of first instance):
The appeal is dismissed with costs, including the costs of two counsel.
________________________________________________________________
JUDGMENT
________________________________________________________________
BRAND JA (NAVSA AND MHLANTLA JJA CONCURRING):
[1] On 3 March 1994 a motor vehicle collision occurred on the N2 highway near Moorreesburg in the Western Cape. The collision tragically claimed the lives of two young parents, Mr Petrus Geyser MacDonald and his wife, Mrs Myra MacDonald. I propose to refer to them as the deceased father and the deceased mother, respectively. They were survived by three children: Sumé, born on 24 July 1984; Petrus, born on 15 February1988 and Lize-Mari, born on 30 March 1992 who are the appellants in this matter.
[2] The respondent is the Road Accident Fund, a statutory insurer, established in terms of the Road Accident Fund Act 56 of 1996 (the Act). The proceedings which eventually led to this appeal commenced when a curator ad litem for the three appellants, who were still minors at the time, instituted an action in the Western Cape High Court against the respondent for the loss of support that they had suffered due to the death of their parents. Because of the lapse of time since the collision, all three children have since attained majority. In consequence they are now cited as appellants in their own names. I propose to refer to them by their first names, or collectively, as the appellants.
[3] The respondent conceded that the collision was caused by the negligence of the insured driver, who collided with the vehicle of the deceased parents and that, in consequence, it was liable for the appellants for whatever loss of support they could establish at the trial. The issues presented to the court a quo for determination, thus turned on the quantum of the appellants’ respective claims. At the end of the proceedings Bozalek J absolved the respondent from the instance, essentially on the basis that the appellants had failed to establish that their reasonable maintenance needs could not be met by the proceeds of the deceased father’s estate. The appeal against that judgment is with the leave of the court a quo.
[4] The issues arising on appeal will best be understood in the light of the background facts that follow, which proved to be largely undisputed. Hence issues of credibility do not arise. The deceased parents were married out of community of property. During the lifetime of the deceased father he owned a farm and derived income from his farming activities as well as from a small transport business. Though the amount of the deceased father’s projected future income was challenged by the respondent, the court a quo determined this amount at the inflation adjusted value of his income for the 1994 financial year which was R218 000. According to the evidence, the deceased mother also generated a small income by way of selling items of pottery as well as milk and meat produced on the farm. The court a quo found, however, that the appellants had failed to establish that the deceased mother had made an independent monetary contribution to their maintenance. Even at this early stage it can be stated, in my view, that these factual findings of the court a quo regarding the financial position of the deceased parents were not and could not be criticised by counsel for the respective parties.
[5] Six months before their untimely death, the deceased parents executed a joint will. The will provided, inter alia, for the event of their simultaneous demise. In which case they bequeathed both their estates to a testamentary trust to be created for the benefit of their children. As to the powers of the trustees, the will authorised them to utilise so much of the income and the capital assets of the trust as they regarded necessary, in the exercise of their absolute discretion, for the maintenance, upbringing, education and other interests of the children. According to the will, the testamentary trust thus created would be dissolved when the youngest child, Liza-Mari, turned 21 which, as we know, will occur on 30 March 2013.
[6] In accordance with the nominations in the joint will, the sister of the deceased father, Mrs Susan van Rensburg, was appointed as the executrix in both estates while the brother of the deceased father, Mr Hermanus MacDonald, and his wife, Marlene (the MacDonalds) were appointed as guardians of the minor children. Shortly after the death of the deceased parents, the appellants moved in with the MacDonalds who looked after them admirably and took responsibility for their upbringing until they each reached the age of independence. Mrs van Rensburg took control of the financial affairs of the estates. Unfortunately, she also died in a motor vehicle accident in March 1997. After her death an attorney, Mr Phillipus de Villiers, was appointed as the executor in the estates of the deceased.
[7] At the time of her death Mrs van Rensburg had not yet registered the trust envisaged by the deceased parents. That was done by De Villiers only in 1997. De Villiers and the MacDonalds were then formally appointed as joint trustees of the trust. In practice, however, De Villiers took sole control of the trust’s financial affairs. Mrs van Rensburg had also not finalised the liquidation and distribution accounts in the estates. De Villiers thus prepared the accounts which were eventually approved by the Master. According to his evidence, the early history of the financial affairs of the trust was unfortunately left largely obscure. What we do know from the evidence of De Villiers and Herman MacDonald, however, is that Mrs van Rensburg paid an amount of R700 per month per child to the MacDonalds. After the appointment of De Villiers he continued with the same practice.But, as the children grew older, their needs increased and a pattern was then established whereby the MacDonalds were regularly reimbursed by De Villiers for the expenses they incurred on behalf of the children. Over the years, so it seems, a system thus developed which satisfied the requirements of the MacDonalds and at the same time served the best interest of the children.
[8] According to De Villiers’ evidence at the trial, the payments thus made from the death of the deceased parents up to the end of February 2010, amounted to R1 979 390, 46. A schedule prepared by De Villiers also shows that during the first years the amounts paid were relatively small. In fact, during the first four years they were limited to R25 200 (calculated on the basis of R700 x 3 x 12) per year. In later years, from about 2001, the amounts paid increased substantially. The explanation for this phenomenon by Herman MacDonald in his evidence was that the larger amounts included college fees and the purchase price of motor cars for the two older children as well as an amount of R70 000 for the wedding of Sumé in November 2009.
[9] The nett assets in the estates of the deceased parents, which eventually devolved to the trust, appear from the final liquidation and distribution accounts prepared by De Villiers. The assets in the estate of the deceased mother essentially derived from the proceeds of two insurance policies which came to about R326 000. Her only other asset was a credit balance of some R9 000 on a credit card account. Her liabilities, on the other hand, were hardly worthy of mention.
[10] The final liquidation and distribution account in the estate of the deceased father showed the following:
(a) The total nett assets in his estate after provision had been made for liabilities and estate duty came to about R2,2 million which included the proceeds of insurance policies in an amount of some R1 683 281.
(b) The non-insurance portion of his estate included his farm, which was reflected at its Land Bank valuation of R378 000; his shares in Orange Rivier Wine Cellars Co-op Ltd, valued at R8 250; and his farm implements and equipment valued at R94 500.
[11] It is plain, however, that at least some of the assets in the deceased father’s estate were substantially undervalued. This is borne out by an agreement of sale entered into during May 1997. In terms of that agreement the farm was sold for R1,2 million; the shares in the Orange River Wine Cellars for R70 000; and the farm implements and equipment for R200 000. Three assets valued in the liquidation and distribution account at less than R500 000 were thus sold for almost R1,5 million. In terms of the deed of sale the purchase price was payable by way of three instalments of R500 000 each. The first instalment became due upon transfer while the second and third instalments were payable one year and two years later respectively. In terms of the sale interest was payable on the last two instalments at the rate of 18 per cent per annum from date of transfer to date of payment. Eventually the farm was only transferred in 1999 so that the instalments of R500 000 each only became due in 1999, 2000 and 2001.
[12] With the benefit of hindsight, it is apparent that the affairs of the trust were extremely well managed by De Villiers and that, acting on the advice of experts, wise investment decisions were made. In the result the assets of the trust grew steadily and towards the end of February 2010 the total portfolio of these investments stood at about R7,4 million. That was over and above the R1,97 million which had been expended over the years for the benefit of the appellants.
[13] Essentially relying on the facts and figures thus far recounted, an actuary, Mr P W Ennis, was instructed on behalf of the appellants, to calculate the quantum of the loss of the support they each suffered. These calculations by Ennis, in turn, formed the basis of their claims in the court a quo. But before I revert to those calculations, I find it convenient to turn to the legal principles involved.
[14] The historical background and the general principles underlying the action of dependants for loss of support by their breadwinners, have by now become well known (see eg Jameson’s Minors v Central South African Railways 1908 TS 575, 602 et seq; Hulley v Cox 1923 AD 234; Evins v Shield Insurance Co Ltd 1980 (2) SA 814 (A); Boberg’s Law of Persons and the Family 2 ed (by Belinda van Heerden, Alfred Cockrell and Raylene Keightley) at 298 et seq; Joubert (ed) The Law of South Africa 2 ed Vol 7 para 88). As I see it, a restatement of these principles can hardly serve any purpose. Hence I focus the spotlight only on those principles that have a direct bearing on this case.
[15] First amongst these relevant principles is that loss of support is confined to actual pecuniary loss. In the first place that means that the dependants cannot claim compensation in the form of a solatium for the grief, the stress and the hurt brought about by the death of a loved one, because these are not capable of being calculated in money. It also means that the dependants are not allowed to profit from the wrongdoing of the defendant. Accordingly, the actual pecuniary loss to which the dependants are entitled, can only be ascertained by a balance of losses and gains, that is by having regard not only to the losses suffered, but also to the pecuniary advantage which may come to the dependants by reason of the breadwinner’s death (see eg Hulley v Cox, supra, 243; Indrani v African Guarantee and Indemnity Co Ltd 1968 (4) SA 606 (D) at 607F-H). This requires one to take into account any income available to the dependants by way of an inheritance from the erstwhile breadwinner (see eg Jameson’s Minors, supra, 603-604). And it matters not, as I see it, whether the income thus available would come to the dependants directly from the deceased breadwinner’s estate or indirectly through the mechanism of a trust. If the nett assets in the estate of a deceased parent, together with the income derived from those assets are therefore sufficient to support the dependants in full, no claim for loss of support as a result of the breadwinner’s death can be sustained (see eg Lambrakis v Santam Ltd 2002 (3) SA 710 (SCA) paras 19 and 20).
[16] The second legal principle relevant to this case derives from statute, more particularly from the Assessment of Damages Act 9 of 1969. The Act is a model of brevity. Its operative provisions are all contained in s 1, which provides:
‘When in an action, the cause of which arose after the commencement of this Act, damages are assessed for loss of support as a result of a person’s death, no insurance money, pension or benefit which has been or will or may be paid as a result of the death shall be taken into account.’
[17] This brings me to the calculation of the quantum of the appellants’ claims by their actuary, Ennis. These calculations were reflected in a number of reports. The hypotheses and factual bases relied upon in these different reports slightly differed in each case. I shall soon return to these differences. But what these reports have in common is that they all start with a calculation of the amount of maintenance the appellants would have received from their deceased parents had they not died. With regard to the deceased mother Ennis’ calculation departed from the assumption that, at the time of the accident, she had earned R12 756 per annum. However, as I have indicated in the course of the factual narrative, the court a quo held this assumption to be unsubstantiated by the evidence. Since I am of the view that this finding is unassailable, we do not have to concern ourselves any further with the appellants’ claims based on the death of the deceased mother. This results from the trite principle that no loss of support is claimable unless it is established that the deceased actually contributed to the support of the plaintiff-dependant.
[18] As to the appellants’ loss suffered through the death of their father, the first premise of Ennis’ calculation was the assumption that he had earned R218 761 per annum when he died and that his earnings would have remained level with adjustments for inflation in later years. He then assumed that Sumé would have remained dependant until 23 – which was an historical fact – while Petrus and Lize-Mari would have become self-supporting at age 21. Finally Ennis assumed that the deceased father’s projected income would have been consumed on the basis of two shares to each parent and one share to each dependant child. After making allowances for certain contingencies, he arrived at the conclusion that because of the death of the deceased father the appellants had been deprived of maintenance in the following amounts:
(a) Sumé R410 581;
(b) Petrus R523 436;
(c) Lize-Mari R779 385.
[19] The court a quo considered these calculations and found some of the assumptions wanting. So, for example, it found the division of two shares for each parent and one share per dependant child, inappropriate in the circumstances. But I have a more fundamental problem with the actuarial approach based on assumptions, hypotheses and contingencies when we know the actual amount that was required for the maintenance of the appellants after the death of their parents. At the time of the trial two of the appellants were already self-supporting while the youngest had almost reached the age of 21. It had also been established that until that stage about R1,97 million was expended for the benefit of the children.
[20] In addition, it would appear that throughout their upbringing all of the appellants’ maintenance needs were met. No evidence was led that the appellants had lacked for anything within reason. On the contrary, it would appear that De Villiers adopted a generous approach to any claim made on their behalf. Indeed, judging by the trust’s contribution of R70 000 towards Sumé’s wedding expenses at a time when she was already self-employed, maintenance needs were liberally construed. There is therefore no reason to think that, but for the death of their father, the appellants would have received more than R1,97 million in maintenance. In the circumstances an elaborate actuarial calculation of the maintenance the appellants would notionally have received seems to amount to an exercise in futility. What is more, it is in conflict with the general principle that ‘where facts are available they are to be preferred to prophecies’. (See Simpson v Jones [1968] 2 All ER 929 (Ch) at 935, referred to with approval in Inter Maritime Management SA v Companhia Portuguesa de Transportes Maritimos EP 1990 (4) SA 850 (A) at 869G-H. See also Road Accident Fund v Monani 2009 (4) SA 327 (SCA) para 9.)
[21] This brings me to the differences between Ennis’ divergent reports, which differences all relate to the monetary benefits the appellants in his view, received from the estate of their deceased father and, more particularly, to the valuation of the assets in his estate. In his first report Ennis relied on the values reflected in the liquidation and distribution account in the deceased father’s estate. On that basis the total capital, nett of estate duty, transferred to the trust amounted to R2 213 511, which included the proceeds of life insurance policies to the value of R1 683 281. On that basis he determined the portion of trust capital not attributable to insurance policies at some 24 per cent. In Ennis’ opinion the Assessment of Damages Act thus requires a deduction of 24 per cent of the maintenance actually received by the appellants, that is R1,97 million, from the amount of maintenance they would have received from their deceased father.
[22] As pointed out in my earlier narrative of the facts, it became apparent, however, that some of the estate assets were substantially undervalued when these assets were sold in May 1997. The disparity was particularly striking with reference to the farm valued for estate purposes at R378 000, to be sold three years later for R1,2 million. Other assets that were sold in May 1997 at markedly higher prices included farm equipment and implements, as well as the shares held by the deceased father in Orange River Wine Cellars Co-op. In his subsequent reports Ennis therefore tried to accommodate these higher values. One of the ways in which he did so was to assume that the Land Bank valuation underlying to the R378 000 was one third less than the true market value of the property. He therefore adjusted the value of the farm upwards by 50 per cent. On this basis the nett asset value of the estate became R2 402 511 of which the non-insurance component changes to about 30 per cent.
[23] Another way in which Ennis tried to provide for the increased value of estate assets was to include the farm at the value of R1,2 million in May 1997, adjusted for inflation back to 3 March 1994, which was the date of the accident. On this basis the value of the farm became R980 985, as a result whereof the nett asset value of the estate grew to R2 816 497, of which the non-insurance component became about 40 per cent. In his final report Ennis again took the value of the farm at R980 985. In addition he adjusted the value of the farming equipment and implements as well as the shares in the Orange River Wine Cellars to reflect the prices at which these assets were eventually sold in May 1997. On this basis he determined the total nett capital received by the trust at R2 983 797 which left the non-insurance component at 43,6 per cent.
[24] The different calculations based on different valuations obviously led to different results in Ennis’ determination of the benefits that came to the appellants through the death of their father. For purposes of further discussion, I will, however, refer to the last report only. From the 43,6 per cent threshold established in that report, Ennis proceeded to calculate that the loss suffered by the appellants as a result of their father’s death, should be reduced by 43,6 per cent of the R1,97 million they actually received from his estate via the trust. On that basis Ennis determined that the appellants’ loss suffered as a result of the death of their father should be reduced by the following amounts:
(a) Sumé R213 447;
(b) Petrus R281 732;
(c) Lize-Mari R444 555
[25] According to Ennis’ calculations the appellants were thus entitled to payment of the following amounts from the respondent:
‘(a) Sumé R 432 191
Less 213 447
R197 134
(b) Petrus R 523 436
Less 281 732
R241 704
(c) Lize-Mari R779 385
Less 444 555
R 334 830’
[26] The appellants’ argument in support of Ennis’ underlying thesis of a percentage split between insurance and non-insurance money received by the trust, proceeded along the following lines. The total monetary benefit received by the appellants from their deceased father’s estate came to R1,97 million. Part of that benefit, however, derived from the proceeds of life insurance policies. In terms of the Assessment of Damages Act, that part ‘shall [not] be taken into account’ in calculating the appellants’ loss of support. We know that both the insurance money and non-insurance money found its way into one communal account of the trust. But it is not possible to determine whether the maintenance payments by the trust came from the proceeds of the insurance policies or from a non-insurance source. For that reason a practical approach to the facts dictates that the same percentage split to maintenance payments should be applied as the percentage split between insurance and non-insurance money that came to the trust from the deceased father’s estate.
[27] The fundamental problem with Ennis’ thesis, as I see it, is that the Assessment of Damages Act does not support the notion of a percentage split on which it relies. As I see it, s 1 of the Act requires that damages for loss of support resulting from a breadwinner’s death must be calculated without reference to the proceeds of insurance policies. In the assessment of benefits resulting from the breadwinner’s death, the proceeds of insurance policies must thus be ignored. The position of both the plaintiff-dependant and the defendant-wrongdoer is the same as if the deceased breadwinner’s estate had received no money from insurance policies at all.
[28] According to the appellants’ argument in support of the percentage split thesis, the Act requires a determination of where the maintenance actually received by the dependants came from. If we know that all the maintenance received by the dependants came from the proceeds of insurance, the Act therefore requires that all maintenance received be ignored. Any other approach, so the appellants’ argument went, will result in the wrongdoer reaping the benefit of insurance policies which is exactly what the Act is intended to prevent.
[29] I agree that the intention of the Act is that a wrongdoer should not benefit from the proceeds of insurance. But I cannot see how the wrongdoer can benefit from the proceeds of an insurance policy if the policy is completely ignored. The inescapable consequence of the appellants’ argument would be that the executor in a deceased breadwinner’s estate, who has both non-insurance and insurance money available, could decide whether the application of the Act should have any effect or none at all. If the executor decides to pay all maintenance from money derived from a non-insurance source, everything is deductable from the maintenance loss. But if the executor should decide to pay all maintenance from insurance money these payments are ignored.
[30] Moreover, the appellants’ argument avoids the question as to what the approach should be in the event that no maintenance had yet been paid when the assessment is made. Should one’s approach then also be that the assessment depends on the executor’s likely decision whether to pay the future maintenance from insurance policies or from non-insurance sources where both are available? I think the answer to the question posed is obvious. One should ignore the insurance policy and enquire whether there is sufficient non-insurance money in the estate to meet the future maintenance needs of the dependants. As I see it one’s approach can be no different if the maintenance had already been paid. The Act requires one to ignore the proceeds of insurance policies. Whether the maintenance actually paid came from an insurance or a non-insurance source is of no consequence.
[31] In accordance with what I see as the correct application of the Act, the first step is thus to ignore the proceeds of insurance policies. The second step that follows is an enquiry which is no different from the one embarked upon in a case such as Lambrakis, supra, where there is no insurance money available. That inquiry is in essence whether the (non-insurance) assets in the deceased estate together with the income derived from those assets would be sufficient to meet the maintenance needs of the dependants (see Lambrakis, supra, paras 19 and 20). In a case such as the present where the maintenance needs of the dependants had already been met, the inquiry is even less complicated. It is simply whether the amount of maintenance we know the dependants had actually received could be produced by the (non-insurance) assets in the estate together with the income generated by those assets. That is the actuarial exercise Ennis should have been asked to perform, but which he was obviously not requested to do.
[32] As appears from what I have said with reference to Ennis’ divergent reports, he was asked to determine the notional market value of the non-insurance assets in the deceased father’s estate as at the date when he died. That, as I see it, was an irrelevant exercise which led to equally irrelevant answers. What we need to know was whether the maintenance actually paid could be covered by the non-insurance part of the trust assets. Far more relevant for that purpose, for example, is the purchase price actually received for those assets instead of their notional market value at some stage in the past. It is therefore again a case of relying on notional calculations based on hypotheses and assumptions rather than available facts. Because Ennis was not asked to determine whether the non-insurance part of the estate was sufficient to pay the amount of R1,97 million disbursed in respect of maintenance, we do not know that exact answer to this question, which lies at the heart of the case.
[33] But even a rough, non-detailed appraisal of the available figures seems to indicate that the non-insurance part of the trust was sufficient to cover the appellants’ maintenance needs of R1,97 million. If I understood the appellants’ counsel correctly, that much was conceded by him and in my view rightly so. We know, for example, that from 1999 until 2001 the trust received an amount of R1,5 million together with interest in an amount of R270 000 from the sale of the farm and other movables. That in itself would prima facie be sufficient to cover the maintenance payments subsequently made. We also know that from the schedules produced on behalf of the appellants, the maintenance they received between 1994 and 1998 was no more than R25 200 per annum and in 1999 it was R35 200. Despite the absence of any direct evidence with regard to the income of the trust during those early years, there is no reason to believe that it did not exceed these moderate amounts. We also know, for example, that the trust is the owner of a flat valued at R650 000 which is occupied by the appellants’ grandmother and from which no income had been received by the trust. Finally, we know that after all maintenance needs of the appellants had been met there is still an amount of about R7,4 million left in the trust. I am not losing sight, of course, of the fact that a large part of both the R1,97 million paid out as maintenance and the remaining R7,4 million originated from insurance policies. Nonetheless, the sheer magnitude of the remaining amount seems to support the conclusion that the non-insurance part of the deceased father’s estate was more then sufficient to provide the maintenance that the appellants actually received. However, in the absence of any exact actuarial calculations, it is perhaps advisable to adopt the more cautious approach followed by the court a quo. That is, to absolve the respondent from the instance rather than to dismiss the appellants’ claims.
[34] In the result the appeal is dismissed with costs, including the costs of two counsel.
_________________
F D J BRAND
JUDGE OF APPEAL
APPEARANCES:
For AppellantS: D J Coetsee
Instructed by: Honey Attorneys
TYGER VALLEY
Correspondents: Honey & Partners
BLOEMFONTEIN
For Respondent: D Jacobs SC
S Mahomed
Instructed by: Z Abdurahman Attorneys
CAPE TOWN
Correspondents: Webbers
BLOEMFONTEIN