Work kills far fewer Australians than in 2003. It disables about as many.

What the indicator is

Burden of disease is measured in disability-adjusted life years, and the scorecard sets out the construction rather than assuming it: The number of Disability-adjusted life years (DALY) lost is the sum of: Years of life lost (YLL), and Years of life lived with disability (YLD). The occupational slice comes from the Australian Institute of Health and Welfare, and the rate is Safe Work Australia’s own construction on top of it. The report says the AIHW study calculates the YLL and YLD due specifically to occupational hazards and exposures, and that We can combine this with ABS employment data to calculate rates of YLL and YLD per 1000 workers.

So this is not a count of injuries or of claims. It is an estimate of how much healthy life Australian work costs, per thousand workers, and it is one of the few measures in the fleet of workplace-safety statistics that reaches back two decades on a consistent basis.

The series

Rates per 1,000 workers, from the supplementary indicators of the Year 6 scorecard, sourced there to the AIHW Burden of Disease 2024 study and the ABS Labour Force Survey of August 2025. The Total column is our addition of the two published figures, which is what the report says a DALY is.

Three things follow from those ten numbers, and the arithmetic is ours on Safe Work Australia’s figures.

The fatal side fell by about 42 per cent. From 3.6 to 2.1 is a reduction of one and a half years of life lost per 1,000 workers.

The disabling side fell by about 4 per cent, from 4.4 to 4.2, and it did not fall smoothly: the 2018 figure of 4.4 is above the 2015 figure of 4.2, so the honest description is Safe Work Australia’s own, that it has changed little. Anyone reading a trend into a wobble of two tenths is reading too much.

So 88 per cent of the total improvement came from the fatal side. Total burden fell from 8.0 to 6.3, a reduction of 1.7, of which 1.5 is years of life lost. And the composition inverted along the way: in 2003 disability accounted for a little over half the burden, and by 2024 it is exactly double the years of life lost.

Why this is the more useful number

Fatalities are what workplace safety is measured by in public. They are counted precisely, reported annually, and they are the thing a regulator can point at. On this series they are also the thing that has genuinely improved.

Our view, built on the figures above: a prevention effort optimised for the countable outcome has moved the countable outcome. The harm that does not kill people, and instead is carried for years, has not shifted in two decades. That is not an argument against the fatality work, which saved the equivalent of one and a half years of life per 1,000 workers and is the clearest public health win in the series. It is an argument that the remaining burden is now mostly of a different kind, and it is the kind that a fatality count cannot see.

The same report supports that reading elsewhere without our help. Its Year 6 measures record that workers with psychological injuries return to work at a markedly lower rate than those with physical injuries, that insurer decision times for mental health condition claims remain more than double the decision time of non-Mental Health condition claims, and that worker experiences of the compensation system have generally deteriorated. Long-duration, hard-to-attribute, hard-to-process harm is the shape of what is left.

What it does not say

It does not say why. A burden-of-disease estimate attributes health loss to exposures using a model, and the model is not a claims dataset: it will include harm that no employer ever heard about and no scheme ever paid for. That is a strength for measuring the real cost and a weakness for anyone wanting to act on a specific hazard.

It also does not tell you which exposures. The AIHW study behind it puts occupational exposures and hazards at 1.6 per cent of Australia’s total burden of disease, which we have written about before, but the scorecard’s supplementary table does not break the occupational slice down further. For duty holders the practical read is narrower than the headline: the number says where the remaining harm sits in kind, not where it sits in your workplace.

Spotted an error? Request a correction and we will check it against the source and log the outcome here.