Why Two Decades of Data Haven’t Moved Construction Suicide Rates

A male construction or mining labourer in Australia is three to four times more likely to die by suicide than a worker in most other industries. That is not a projection or a modelled estimate. It is what the record of more than 20,000 deaths across two decades now confirms.

The finding comes from a study published in BMJ Public Health on 18 September 2026, described by its authors as the first investigation of its kind into occupation-specific suicide patterns in Australia. It spans 43 occupational groups across the years 2001 to 2022, and it sits inside a sobering national picture: 3,311 Australians died by suicide in 2023, of whom 75.3% were male.

This is not an outlier reading from a single bad year. It is a structural signal that has held steady across two full decades.

After reading this, you will have a clear picture of what the evidence actually says about why these rates persist, what the best-evidenced prevention programme has and has not demonstrated, and what structural changes researchers argue are the minimum required for the numbers to move. Treat it as practical intelligence on an embedded operational and regulatory risk, not a welfare summary.

What the BMJ study actually found, and why two decades of persistence is the real headline

Start with the scale, because it is what gives the finding its weight. The study, “Suicide across occupational groups in Australia, 2001-2022: a retrospective mortality study” (DOI: 10.1136/bmjph-2025-003662), analysed more than 20,000 suicide deaths among employed Australians aged 15 to 64. It was led by Humaira Maheen at the University of Melbourne, with RMIT Associate Professor Tania King as senior author.

The primary result is stark. Male construction and mining labourers carry suicide rates three to four times higher than most other occupational groups, placing them at the top of a hierarchy that also includes machine and plant operators, agricultural and forestry workers, and automotive and engineering trades workers.

The persistence is the sharper finding. Only six of the 43 occupational groups recorded meaningful reductions over the two decades, and none of them were the highest-risk groups.

Those six were:

• Business, HR and marketing professionals
• Automotive and engineering trades workers
• Other technicians and trades workers
• Sales representatives and agents
• Store persons
• Cleaners and laundry workers

The relative positioning across the workforce sorts roughly into tiers.

Associate Professor King’s reading of the data is the part that should hold an operator’s attention.

Suicide risk is not randomly distributed. It follows identifiable patterns when occupational specialisation, gender, and certain working conditions intersect, and the absence of progress across two decades points to a problem that is at least partly systemic rather than individual.

Here is what that tells you. If the rates have not shifted in twenty years, across two economic cycles and multiple regulatory regimes and a large national mental health spend, this is not a problem that will self-correct. For anyone with operational or investment exposure to these sectors, it reads as embedded systemic risk, and the study is the first rigorous map of exactly where that risk sits.

Why these industries produce this outcome, the mechanisms researchers have identified

The risk does not come from one stressor. It comes from several that compound, and understanding the sequence matters more than cataloguing the parts.

Researchers treat the drivers as additive and interacting rather than independent. The cluster looks like this:

• Cultural norms of toughness that suppress help-seeking and disclosure of distress
• Hazardous, physically demanding work with high exposure to injury and pain
• Long, irregular or remote-site rosters that isolate workers from family and community
• Employment insecurity and subcontracting that raise financial stress and cut bargaining power
• Low job control, meaning workers have little say over the conditions of their work

Take the masculinity dimension first, because it is easy to dismiss as a soft factor. It is not treated that way in the research. Programme designers target it directly precisely because it is viewed as culturally embedded in these workplaces, a structural feature rather than an individual attitude an awareness poster can fix.

Then layer the fly-in fly-out reality on top: weeks away from home, disrupted family life, isolation on remote sites. Add the subcontracting model, where casualised, short-term work means income insecurity and little ability to push back. None of these is catastrophic alone. Combined, and sustained over years, they erode mental health progressively.

Researchers differ in emphasis rather than substance. Some foreground workplace culture and masculinity norms; others emphasise FIFO conditions, injury exposure, and the financial stress of precarious contracting. The consensus treats them as a cluster, not competing theories.

For operators and investors, the implication is uncomfortable but clear. If the drivers are built into how work is structured and contracted, an employee assistance programme or a wellness app on its own is unlikely to move the risk profile. The stressors are not sitting in the individual worker’s resilience. They are sitting in the work.

The same pattern internationally

The Australian result is not an Australian anomaly. National occupational mortality data in the United States has consistently placed construction and extraction workers among the highest-risk groups, with the same drivers cited: job insecurity, seasonal work, injury, and cultures that discourage help-seeking. The Construction Industry Alliance for Suicide Prevention in the US runs a programmatic response built on similar logic to Australia’s, combining awareness training, leadership engagement and peer support.

UK official statistics tell the same story, showing elevated suicide rates among low-skilled male labourers, construction workers and tradespeople.

That cross-national consistency is itself evidence. When the same occupational groups show elevated rates across three comparable economies, the cause is far more likely to be a structural feature of the work than anything unique to a single country’s culture.

What MATES in Construction demonstrates, and where the evidence runs out

Founded in 2007, MATES in Construction is the most widely implemented workplace suicide prevention programme in the Australian sector, and it has genuine achievements to point to. It targets four domains:

• Suicide prevention literacy
• Help-seeking intentions
• Helping behaviours
• Stigma reduction

The evaluation evidence on those domains is favourable. A 2023 review in Health Promotion International concluded the programme demonstrably improves literacy, help-seeking and stigma reduction among construction workers. On the harder question of mortality, two studies stand out.

Doran et al. (2016), a retrospective ecological and economic evaluation, found a 10% reduction in relative risk (RRR of 0.9) of suicide behaviour among Queensland construction workers in the five years after implementation compared with the five years before.

Maheen et al. (2022) reported a sharper divergence.

Between 2001 and 2019, age-standardised suicide rates fell by an average of 3% per year among construction workers, against 1.5% per year for other employed men. The difference was statistically significant.

That is the strongest quantitative signal available linking the programme to reduced mortality. It is also where the evidence ceiling arrives.

Both studies use ecological, retrospective designs. They can show an association between the programme’s rollout and falling rates, but they cannot cleanly separate that from economic cycles, changes in coronial classification practices, or the many concurrent national mental health initiatives running over the same years. The 2023 review made the same point plainly: the causal evidence base remains limited, with very few controlled evaluations and no experimental studies.

Do not let the strong evidence for literacy and stigma gains bleed into a claim of proven mortality reduction. They are different things, and the research keeps them separate.

The evidence is now moving toward firmer ground. RMIT’s MatesMonitor project is running a prospective, longitudinal evaluation of the service, and a protocol for MATES in Manufacturing plans a cluster randomised controlled design across up to 12 worksites, a genuine methodological step toward causal proof.

For a procurement or due-diligence decision, the honest read is this. MATES-type programmes are the best-evidenced intervention on offer and worth implementing. But knowing where the evidence is strong and where it is genuinely uncertain is what separates substantive risk reduction from compliance optics, and these programmes cannot substitute for changing the conditions that generate the risk.

The structural reform argument, what researchers say must change at industry level

Follow the logic through and the conclusion arrives on its own. Two decades of data show no meaningful improvement. The mechanisms are embedded in how work is organised and contracted. The best-evidenced programme lifts literacy and help-seeking but has not been shown to reshape mortality on its own. Put those together and individual-level interventions cannot be the whole strategy.

Researchers and MATES-linked practitioners identify four reform domains, and they are distinct priorities rather than interchangeable suggestions.

1. Employment security and contracting. Moving away from highly casualised, short-term arrangements toward more stable relationships, reducing the financial insecurity repeatedly linked to distress in blue-collar workforces.
2. Working hours, rosters and fatigue. Limiting excessive overtime, making rosters more predictable and family-friendly, and treating FIFO isolation and fatigue as psychosocial hazards rather than productivity questions.
3. Psychosocial hazard regulation. Treating high job demands, low autonomy, poor supervision and bullying or harassment as regulated hazards employers must actively control and monitor, integrated into safety systems like any other serious risk.
4. Industry standards and accountability. Embedding mental health and suicide prevention requirements into industry codes, contracts and procurement, and making senior leaders and principal contractors accountable rather than leaving it to individual supervisors.

Associate Professor King frames peer-led programmes such as MATES as one component of the answer rather than the whole of it, while calling for broader systemic change directed at the occupational groups carrying the heaviest burden.

The counterargument and why occupational researchers push back

There is a legitimate opposing view. Some clinicians and policymakers frame suicide primarily as a medical issue, best addressed through better access to general mental health services and individual clinical care rather than occupational restructuring.

Occupational health researchers push back on a specific point. That framing struggles to explain why construction and mining workers in particular show persistently elevated rates, and it does not account for why broad national mental health investment has failed to move those rates over twenty years. If the risk were purely individual and clinical, the concentration in specific sectors would be far harder to explain.

For operators and investors, the direction of regulatory travel is toward mandatory psychosocial hazard management. That means a company treating workforce mental health as a compliance checkbox rather than an operational system is increasingly exposed, both to regulators and to the social licence conditions that shape extractive approvals in Australia. The reform agenda touches procurement, contractor management and duty of care directly.

What a two-decade flat line demands from the industry now

Pull the threads together. The BMJ study confirms construction and mining labourers carry a suicide burden three to four times the norm. That burden has not improved across two decades. The mechanisms driving it are structural. And the best-evidenced programme delivers real gains in literacy and help-seeking but cannot, on its own, be shown to substitute for systemic reform.

Three pressures are now converging. The first is evidentiary: a study of more than 20,000 deaths across 43 occupational groups over 2001 to 2022 is not easily dismissed, and it sits within a national toll of 3,311 suicide deaths in 2023 (ABS revised figure, 7 May 2026). The second is regulatory, with Australian workplaces moving toward mandatory psychosocial hazard management. The third is the tightening social licence scrutiny on extractive industries’ duty-of-care obligations.

The next generation of data will be harder to argue with still. MatesMonitor’s prospective design and the MATES in Manufacturing randomised trial protocol are built to deliver the causal evidence the ecological studies could not.

Here is the practical implication. The question the evidence now poses is no longer whether this is a systemic problem. It is what systemic response you are prepared to support. Waiting for more proof before treating this as a managed operational risk is not a neutral decision to await clarity; it is a choice to accept the current exposure.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions.

If you or someone you know is struggling, support is available in Australia through Lifeline on 13 11 14.