DEATH BENEFITS: Editorials: October 2019 / January 2020


SCA gets it

Court’s interpretation of “dependants”, Karin Lehmann* believes,
creates greater confusion for trustees.

In its recent FundsatWork v Guarnieri judgment, the Supreme Court of Appeal misconstrued both the definition of “dependant” as well as s37C of the Pension Funds Act.

The facts were briefly that the fund’s trustees had allocated the greater share of a member’s death benefit to his aged mother. This was to the detriment of his estranged widow and children.

Unbeknown to the trustees, the mother had died four days before they exercised their discretion. By the time they learned of the mother’s death, the money had been disbursed to her and was no longer recoverable.

Although the equitability of the trustees’ allocation is questionable, the trustees were understandably reluctant to comply with the High Court order that they pay the equivalent sum to the deceased member’s widow and children; hence their appeal to the SCA.

The SCA’s view was that the validity of the trustees’ allocation turned on whether the mother was a dependant at the operative time for determining eligibility. This, it held, was when the trustees actually exercise their discretion rather than at the date of the member’s death.

The mother was thus not an eligible dependant. The SCA went even further, cryptically stating that the dependant “must still be a beneficiary” when payment is made.

Almost every aspect of the SCA’s judgment is troubling. Incomprehensibly it suggested that it will not impose an undue “practical burden” on trustees to ensure that facts — painstakingly collected over a period of months — are still true on the date of allocation and payment.

Life is not static. Circumstances change. If facts were instantly verifiable, trustees would not by law have been given 12 months to complete their investigation.

Of greater concern is the SCA’s understanding of “dependants” under s37C(1)(a). Rather than adopting the obvious interpretation, which is that this paragraph pertains to the member’s inter vivos dependants (in accordance with the Adjudicators’ interpretation), the SCA held that it applies to persons owed a posthumous duty of support. These are principally needy spouses and children.

The SCA concluded that the legislature’s “careful and deliberate” use of the present and past tenses in the definition indicated that the paragraph was aimed at protecting those to whom the member owes a present duty of support and that “present” is the date on which the trustees exercise their discretion.

In other words, use of the word “is” in the paragraph must be read literally by trustees. They must then ask themselves: “As we gather here today, who ‘is’ owed an existing duty of support i.e. one that has not been terminated by the member’s death?”

The SCA’s interpretation is flawed. Spouses and children are already, in every case, statutory dependants. It matters not whether they are amongst a member’s legal dependants or financial dependants or neither. They are, without more, dependants.

If the paragraph is aimed at spouses and children, it is redundant. Moreover, the current definition was adopted in 1989. This was before the 1990 Maintenance of Surviving Spouses Act created posthumous duties of support towards spouses.

The SCA’s interpretation of s37C(1)(a) is also incompatible with s37C(1)(c). The SCA accepts that a legal duty may arise after the member’s death, while concluding that an existing duty is extinguished by the member’s death.

A parent, sibling, grandchild or grandparent towards whom a duty is not yet owed — while the member is alive — may thus be an eligible future dependant. A parent and others, to whom the member whilst alive already owes a duty of support, automatically becomes ineligible on the member’s death.

The SCA thus includes contingent legal dependants but excludes existing legal dependants. This is at odds with its own observation that courts prefer “facts” to “prophecies”.

The judgment contains further errors. It equates a nominated beneficiary with a nominee, and thus misunderstands when s 37C(1)(a), (b) or (bA) applies. It conflates equitability with financial need.

It believes that if the relevant date for determining eligibility is the date of death, the facts relevant to an equitable allocation are similarly those that existed at the date of death. And, most extraordinarily, it states that s 37C does not “entirely override” a member’s wishes.

If the SCA can get so much so wrong, it is beyond time that s 37C be revised. ν

*Lehmann is a UCT law lecturer and a trustee actively involved in making s37C decisions.

Lehmann . . . peculiar decision


The SCA thus includes contingent legal dependants but excludes existing legal dependants. This is at odds with its own observation that courts prefer “facts” to “prophecies”.