If you read enough actuarial commentary and legal judgments on loss of earnings claims in South Africa, a pattern emerges quickly: there is no single, universally precise way to arrive at a number. Actuaries, attorneys, and the courts themselves describe a process that is deliberately approximate at several points — and they say so openly. That matters for anyone using (or building) a quick estimation tool, because it means a simplified calculator isn't cutting a corner that “real” methodology avoids. The real methodology has the same corner built into it.
The actuarial starting point
The standard actuarial approach, as described by South African forensic actuaries, works in two halves. First, the actuary calculates pre-morbid earnings — what the claimant would have earned had the accident not happened, projected forward using assumed salary growth to retirement. Second, they calculate post-morbid earnings — what the claimant can still be expected to earn in their injured state, if anything.
Both streams are converted to a present value using a life table (in most reports, Robert Koch's tables) and a net discount rate, and the loss is the difference between the two, adjusted for contingencies.
Robert Koch, whose Quantum Yearbook underpins most of this work, is explicit that the actuary's role is not to hand down a verdict. As he's put it, the actuary functions as “an expert calculator, economist and statistician” assisting the court — not a valuator with final say. The number an actuary produces is a structured estimate built on stated assumptions, not a measured fact.
What the courts actually do with that number
This is where it gets interesting for anyone assuming actuarial output is treated as gospel. In the frequently-cited case Southern Insurance Association Ltd v Bailey, Nicholas JA drew a sharp line between two ways a court can arrive at a damages figure: a judge's “round estimate… entirely a matter of guesswork, a blind plunge into the unknown,” versus “an assessment, by way of mathematical calculations, on the basis of assumptions resting on the evidence.” He favoured the latter wherever the evidence supports it, but — and this is the important part — South African courts have never treated the actuarial figure as binding.
As legal academic L Steynberg documents in her 2011 Potchefstroom Electronic Law Journal analysis of the case law, “the court necessarily exercises a wide discretion when it assesses the quantum of damages due to loss of earning capacity” (per Zulman AJ in Road Accident Fund v Guedes), and the actuarial calculation is repeatedly described as “the starting point,” not the destination.
Even earlier, in Hulley v Cox (1923), the court chose a rough estimate over a full annuity calculation specifically because it wanted more flexibility than the mathematical method allowed — and Innes CJ noted it's “desirable to test the result of an actuarial calculation by a consideration of the general equities of the case.” A century of case law later, that's still the operating principle: run the numbers, then sense-check them against fairness.
Contingencies: the part everyone agrees is a judgment call
If any single variable in a loss-of-earnings calculation admits it isn't precise, it's the contingency deduction — the percentage knocked off (or occasionally added to) the capitalised figure to account for things like unemployment risk, illness, early death, or a shortened working life. The Road Accident Fund's own long-standing convention applies “normal” contingencies of roughly 5% for past loss and 15% for future loss, but that's a rule of thumb, not a derived figure. Koch's own guideline for general contingencies is a sliding scale of “half a percent for every year to retirement age” — again, a heuristic, not a calculation.
Actual deductions applied by courts range enormously. Steynberg's review of the case law turns up specific contingency deductions from 0% right up past 70% depending on individual facts — 70% for a remarriage prospect in Milns v Protea Assurance (on top of a further 22% general deduction), 50% in Shield Insurance Co Ltd v Booysen, 70% again in Trimmel v Williams.
In Road Accident Fund v Guedes, the actuary's original 10%/30% split was revised on appeal to 20%/30%, which alone moved the award from R3.12 million to R2,323,633 — a swing entirely attributable to a contingency judgment call, with no new facts about the claimant's earnings at all.
None of this is presented in the literature as a flaw to be embarrassed about. It's presented as how the exercise works: contingencies compress a wide range of unknowable future circumstances into a single adjustable percentage, and reasonable, qualified people can and do disagree on the right number for the same claimant.
What this means for a simplified calculator
Put the pieces together and the picture is consistent across every source, whether written by an actuary, an attorney, or a law journal: even a full actuarial report rests on projected (not certain) earnings growth, a chosen (not measured) discount rate, a selected (not universal) life table, and a contingency percentage that is openly acknowledged as a judgment call rather than a computed value — one the presiding court can, and regularly does, overrule.
Given that, a tool that simplifies further — using a single effective tax rate instead of month-by-month tax iteration, or standard contingency defaults instead of a bespoke industrial psychologist's report — isn't introducing a new kind of imprecision into an otherwise exact process. It's applying the same category of simplification the field already relies on, just earlier in the process and more transparently.
The honest position, and the one this methodology follows, is to be upfront that the output is an indicative estimate for preliminary assessment, not a final court figure — which is precisely the caveat the courts themselves apply to full actuarial reports before they even get to judgment.
A simplified calculator earns trust not by pretending to a precision the underlying discipline doesn't claim for itself, but by being clear about which assumptions it's making — and that's a standard the case law and the actuarial literature both set for it. Try it on the calculator.
Sources
- RAF Loss of Earnings Calculation — Actuary Consulting
- How Actuaries in South Africa Ensure Fair Compensation — Actuary Consulting
- Steynberg L, “Fair Mathematics in Assessing Delictual Damages” 2011 PER 9 — SAFLII
- RAF Loss of Earnings Claims — DSC Attorneys
- Quantifying Damages Claims in South Africa — Lexology
- Calculation of Future Loss of Income — Webbers Attorneys
- Financial Loss Calculations — SNG ARGEN Actuarial Solutions