From $146,000 to $288,000 in five years — the workers’ comp line your CFO can’t ignore

It usually surfaces in a finance review, not a safety one. The workers' comp premium has climbed again. When someone breaks down the claims behind it, the shape has changed: the expensive, slow-closing files are no longer the back injuries and the falls — they are psychological. They stay open for months. And when the CFO asks, plainly, what is driving the line, the answer comes back in the language of wellbeing and engagement rather than in the language of cost, duration and probability. So the question doesn't actually get answered. It gets deferred to next quarter, when the premium notice will be larger.

That gap — between a cost the CFO can see and an explanation the organisation can put in finance terms — is the real problem. It is not a mental health problem. It is a risk-measurement problem, and it is getting more expensive every renewal cycle.

Why has the expensive claim changed shape?

Start with the number that should be on the CFO's desk. The average cost of a psychological injury claim in NSW rose from $146,000 in 2019-20 to $288,542 in 2024-25 — close to a doubling in five years (Insurance Council of Australia, submission to NSW Parliament, 2025). Over the same broad period, psychological claims grew 30% in NSW between FY2018-19 and FY2022-23, against 11% growth for physical claims (SafeWork NSW, Psychological Health and Safety Strategy 2024-2026). The volume is shifting and the unit cost is climbing at the same time. That is the worst combination a ledger can carry.

The duration is where it stops being abstract. Serious mental health claims run a median 35.7 weeks of time lost, against 7.4 weeks for all other serious claims — roughly five times longer — and median compensation sits at $67,400 versus $16,300, around four times higher (Safe Work Australia, cited by SafetySure, 2025). A psychological claim is not a more emotional version of a physical one. It is a structurally more expensive event: longer off work, harder to close, more disruptive to the team left covering the gap.

And the premium follows. The NSW Treasurer's Workers Compensation Ministerial Statement of 18 March 2025 put it bluntly: premiums are predicted to rise 36% over three years without intervention. "Without intervention" is the operative phrase. The forecast assumes organisations keep doing what they are doing now.

Why doesn't your safety dashboard see it coming?

Here is the part that frustrates good operators. You have a mature safety function. You track lost-time injury frequency, near-misses, hazard reports, audit closure rates. None of it moved before the psychological claims started arriving. The dashboard was green.

That is not a failure of the dashboard. It is measuring the wrong layer. Physical-safety leading indicators predict physical incidents — slips, plant, manual handling. Psychological injury is driven by a different set of hazards: unmanageable workload, low job control, poor support from a direct supervisor, role conflict, exposure to aggression, bullying. These are the psychosocial hazards that an employer now has a positive duty to manage under WHS law — the same hierarchy of controls, the same "reasonably practicable" test, the same officer due diligence obligation that already applies to physical risk.

The leading indicators for those hazards do not live in an enterprise dashboard. They live at crew and team level — in the relationship between a supervisor and the fifteen people who actually report to them. That is not a metaphor. Human social capacity tops out around Dunbar's number, the roughly 150-relationship limit; the unit where someone actually notices that a colleague has gone quiet, stopped contributing, started absorbing an unreasonable load, is far smaller than that. The signal is there weeks before it becomes a claim. Almost no organisation collects it, because the infrastructure to make it visible sits in the gap between the safety system above and the EAP below — and nobody owns the gap.

Doesn't the EAP already cover this?

The EAP is necessary. It is also downstream by design. It activates after harm — someone is already struggling enough to pick up the phone. By the time the EAP is in use, the psychosocial hazard that caused the distress has usually been operating for months, and the claim, if it comes, is often already forming. Counting EAP utilisation as your mental health strategy is like counting ambulance call-outs as your road-safety strategy. It tells you about the bottom of the cliff. It tells you nothing about the top.

This is why the spend so often fails the ROI test. KPMG and Mental Health Australia put the return at roughly $4.30 for every $1 invested in workplace mental health (2022). That figure is real — but it is an upstream figure. It is earned by prevention: by identifying the hazard, controlling it, and routing people to support earlier and more appropriately. A dollar spent purely on reactive support — more EAP sessions, another resilience workshop after the fact — does not return $4.30. It returns a smaller, slower number, because it is buying treatment, not prevention. The CFO is right to be sceptical of the spend they can currently see. It is mostly downstream.

What would actually move the line?

The same discipline you already apply to physical risk, applied to the psychosocial kind. A psychosocial safety framework treats workload, control, support and conduct as hazards on a risk register — assessed, controlled, reviewed — rather than as a culture survey filed once a year. The point is to make the invisible risk visible at the level where it is actually generated, so that a supervisor has a structured way to notice and act, instead of becoming an accidental counsellor improvising on instinct.

Done properly, this does two things at once, and they are not in tension. It reduces the need for the EAP — fewer hazards left uncontrolled means fewer escalations into claims. And it increases appropriate uptake — people are routed to support earlier, before the situation has hardened into a 35-week file. Lower claim volume and better-targeted support, from the same intervention. That is what the $4.30 looks like when the dollar lands in the right place.

How do you put this in front of a CFO?

Stop translating it into wellbeing and start presenting it as exposure. The CFO is not resisting mental health; they are resisting a cost they cannot model. Give them three numbers they can model: the claim-cost trajectory ($146,000 to $288,542 in five years), the premium forecast (36% over three years without intervention), and the return on upstream spend ($4.30 per dollar). Then show them the one thing that is currently absent from every board pack — a measured view of where the psychosocial hazard actually sits across the sites and teams, the same way you can already show them where the physical risk sits.

That measured view is what belongs in the gap: a way of managing psychosocial risk that sits pre-EAP, upstream and structured. It is what turns "HR can't answer it in finance language" into a line on the risk register with an owner, a control and a closure date. The CFO doesn't need to be persuaded that mental health matters. They need to be shown that it is now a measurable, controllable, reasonably-practicable risk — and that the cheapest version of managing it is the one that happens before the claim opens.