in in f S

Psychological injury is the most expensive kind of workplace harm to leave alone, and the hardest to see coming, because the people carrying the most load are usually the least likely to say so.

Every figure below is attributed. Where a number is a model rather than an observed result, it says so.

What It Costs

$288,542

Average cost of a psychological injury claim in New South Wales.

NSW Treasurer, ministerial statement, 18 March 2025

12 per cent of claims, 38 per cent of costs

Psychological injury makes up 12 per cent of claims in the New South Wales scheme and 38 per cent of what the scheme pays.

NSW Treasurer, ministerial statement, 18 March 2025

$6 billion

Estimated annual cost of psychological health problems attributable to low psychosocial safety climate in Australian workplaces.

Source: Becher and Dollard 2016. Published by Safe Work Australia.

$2.30 back per $1

Estimated return on investment from workplace mental health interventions.

Source: PwC 2014. For beyondblue. This is an economic model, not an observed result.

Where the Harm Comes From

73 per Cent

Of serious mental stress claims are attributed to harassment and workplace bullying at 33.2 per cent, work pressure at 24.2 per cent, and exposure to violence and harassment at 15.7 per cent. All three are matters of how work is designed and managed.

Mental health conditions were 12.0 per cent of all serious workers compensation claims in 2023 to 2024, 17,600 claims, up 161 per cent in ten years. Median time lost was 35.7 weeks against 7.4 for all serious claims. Median compensation paid was $67,400 against $16,300.

For women, mental health conditions are 17.2 per cent of serious claims. For men, 8.2 per cent.

Source: Safe Work Australia 2025. Key Work Health and Safety Statistics Australia, 2023 to 2024 preliminary claims data. Time lost and compensation are 2022 to 2023. Mental stress sub-category figures exclude Victoria, which codes these claims differently.

Who Carries It

15 to 20 per cent

Of the population is neurodivergent. Most are undiagnosed or undisclosed, so no headcount tells you who is carrying the load.

Doyle, 2020. British Medical Bulletin, 135, 108 to 125. The figure is Doyle's estimate from combined prevalence rates, not a measured result

49 per cent against 32 per cent

Of neurodivergent workers reported one or more harmful workplace events in the past twelve months, against a matched comparison group.

Source: SuperFriend and Specialisterne 2025. From 2024 survey data, 865 neurodivergent workers matched by age and gender.

32 per cent against 17 per cent

Of neurodivergent workers experienced severe psychological distress in the previous four weeks, against the same matched comparison group.

Source: SuperFriend and Specialisterne 2025. From 2024 survey data.

Most adjustments cost nothing

61 per cent of workplace accommodations cost nothing at all. 33 per cent had a one-off cost, median USD $300. 6 per cent had an ongoing cost, median USD $2,400 a year.

Source: Job Accommodation Network 2025. September 2025 update. United States figures in United States dollars, included because no equivalent Australian dataset exists.

Schools

The figures used on the Belonging in Education page. Each entry names the sample, because most of what exists on autistic students in Australian schools is parent report, recruited through advocacy networks.

Six full days every four weeks

Autistic children missed on average six full days of school in a 20-day period, about four times the general rate. School refusal or reluctance accounted for 57.2 per cent of full-day absences. No child missed a full day through truancy.

Source: Adams 2022. Child and parental mental health as correlates of school non-attendance and school refusal in children on the autism spectrum. Journal of Autism and Developmental Disorders, 52(8), 3353 to 3365. doi 10.1007/s10803-021-05211-5. Parent report, 106 children, recruited online, pre-COVID. Families with no attendance problems were likely under-represented, as the author notes.

34 per cent changed primary school

Of families of autistic students, 34 per cent had changed primary school at least once because the school could not support their child. Less than half received adjustments to curriculum and assessment or regular communication and meetings. Less than one in eight said the school provided autism professional development to staff.

Source: Thom-Jones, Gordon and Mizzi 2025. 'I shouldn't have to educate their school about what autism is'. Autism, 29(10), 2451 to 2462. doi 10.1177/13623613251338908. Parent surveys 2020 (N 2212) and 2021 (N 528), recruited through advocacy mailing lists and social media, Victoria-weighted. Funded by Amaze.

47 per cent changed schools

Of parents of autistic children, 47 per cent had changed schools at least once, 35 per cent had been discouraged from enrolling or refused, and 43 per cent reported school refusal at least once a week.

Source: Autism Awareness Australia 2024. Understanding the barriers: the educational experiences of children with autism. National Autism Education Survey, February 2024. Self-selected parent survey, all states and territories; the report does not state the number of respondents.

Senate inquiry, 2023

The committee found school refusal "most prevalent among students with disability (particularly neurodivergent students)", and that almost all parents who gave evidence had already sought help from paediatricians, psychologists and mental health services. Dr Greg Elliott of Catholic Schools Parramatta Diocese told it attendance problems are often "a proxy for students who aren't getting the adjustments they need at school".

Source: Senate Education and Employment References Committee 2023. The national trend of school refusal and related matters. Chapter 2, paragraphs 2.12, 2.33 and 2.56. Primary parliamentary document. Note the 43 per cent figure at paragraph 2.33 is a study cited by a submitter, Triple P International, Submission 140, not Australian data.

What These Numbers Do Not Tell You

These are national and state figures. They describe the size of the problem across Australia. They are not a prediction about any one organisation, and they are associations rather than proof that one intervention caused one result.

The New South Wales figures above are here for one reason: the cheapest claim is the one that never opens. They are an argument for prevention, not for narrowing what injured workers are owed.

The honest way to answer the cost question is to measure your own baseline first, then compare it at six and twelve months.

The longer writing behind these numbers is on Substack. Read the writing on Substack.