Why yoga won’t fix psychosocial hazards: what leaders should do instead

Wellness programs can be good for morale, recovery and culture, but they are not, by themselves, controls for psychosocial hazards. A yoga class, morning tea, mindfulness app or fruit bowl may support individual wellbeing, but it does not remove the source of harm if the real hazard is unmanageable workload, poor role clarity, bullying, traumatic exposure, fatigue, understaffing, low job control or unsafe performance management.

Safe Work Australia draws this distinction directly. WHS is about preventing workers being harmed at work, while wellbeing initiatives such as meditation, healthy eating and exercise may have positive impacts but “do not protect workers from harm from psychosocial or physical hazards” if they do not prevent harm arising from the work itself.

This matters legally as well as practically. The psychosocial hazards Code of Practice is an approved code under the WHS Act, and approved codes can be used in court as evidence of what is known about a hazard, risk or control, and what may be reasonably practicable. WHS Regulation also requires duty holders to apply the hierarchy of controls: eliminate psychosocial risks where reasonably practicable, and otherwise minimise them, starting with higher-order controls such as redesigning work before relying on lower-order controls such as training or coping strategies.

The NSW Police example: a warning about support without prevention

The NSW Police Force example comes from the Audit Office of NSW’s 2025 performance audit into police psychological wellbeing. The audit did not simply say “wellness is bad.” Its finding was sharper: NSW Police had expanded wellbeing initiatives and support services, but it was not adequately identifying and addressing the root causes of psychological injury.

The audit found that NSW Police spent about $34 million on proactive wellbeing services between 2020 and 2025, and about $60 million operating its Health, Safety and Wellbeing Command, while psychological injury costs from July 2019 to June 2024 were about $1.75 billion. The audit also found psychological injury claims rose from 790 in 2019–20 to 1,208 in 2023–24, and psychological claims accounted for 74% of workers compensation costs.

Most importantly, the audit concluded that the wellbeing initiatives were mainly providing counselling, support and debriefing after traumatic incidents, but did not adequately address other psychological risk factors such as fatigue, role overload and burnout. The audit recommended practical controls such as better workforce allocation to match workload, improved incident systems that capture causal factors, investigation of role overload and burnout, and evaluation of whether wellbeing spending is actually linked to evidence-based psychological risk factors.

That is the central lesson for leaders: support services are not the same as risk controls. They may help injured or distressed workers, but they do not prove the organisation has controlled the hazards that are making people unwell.

Why “wellness” is not enough

A workplace wellness program usually targets the individual: exercise more, eat better, meditate, attend a resilience session, call the EAP, join a social event. These may be useful benefits, but they do not change the design or management of work.

The research evidence is also sobering. A large JAMA cluster-randomised trial of more than 32,000 employees found that a workplace wellness program increased some self-reported health behaviours, such as exercise and weight management, but produced no significant differences in clinical markers, health spending, absenteeism, tenure or job performance after 18 months. A separate two-year randomised trial at the University of Illinois similarly found no significant effects on measured physical health outcomes, medical diagnoses or medical use, although it did improve some beliefs and primary care behaviours.

These studies are not saying “never run wellness programs.” They are saying leaders should temper expectations. Wellness programs are not a substitute for controlling psychosocial hazards at the source.

What psychosocial hazards actually are

Psychosocial hazards are not vague “stress” issues. They are identifiable features of work that can cause psychological or physical harm. Safe Work Australia lists common psychosocial hazards including high job demands, low job control, poor support, lack of role clarity, poor organisational change management, inadequate recognition, poor organisational justice, traumatic events or material, remote or isolated work, poor physical environments, violence, bullying, harassment, conflict and poor workplace relationships.

The national data shows why this is now a major governance issue. In 2023–24, Safe Work Australia reported 17,600 serious workers compensation claims for mental health conditions, representing 12% of serious claims. Mental health claims had a median time lost of 35.7 weeks, almost five times the median for all serious claims, and median compensation of $67,400, more than four times the all-claims median.

The legal trend: courts and regulators look for real controls

Recent Australian prosecutions reinforce the same point: regulators and courts are looking at whether the employer identified psychosocial hazards, assessed risks and implemented controls.

In Victoria, Court Services Victoria was convicted and fined over a toxic workplace culture at the Coroners Court of Victoria. WorkSafe Victoria said workers were exposed to traumatic materials, role conflict, high workloads, poor workplace relationships and inappropriate behaviours, and that the employer failed to conduct adequate processes to identify or assess risks to psychological health.

In the Commonwealth jurisdiction, the Department of Defence was convicted in December 2025 for failing to manage psychosocial risks connected with a worker’s death. Comcare said available controls included supervisor training to understand how a performance management process could become a psychosocial hazard, identify risks for workers subject to that process, and suspend or modify the process where needed.

The practical message is that a court is unlikely to be persuaded by “we offered yoga” if the actual allegation is that the organisation failed to manage workload, bullying, traumatic exposure, unsafe rostering, poor supervision, role conflict or unreasonable performance processes.

What interventions actually work?

The strongest interventions are those that change the work, not those that ask workers to become more resilient to harmful work.

The World Health Organization recommends organisational interventions that directly target working conditions and working environments, such as flexible working arrangements and frameworks to address violence and harassment. WHO also recommends manager training, worker mental health literacy and individual stress-management interventions, but these sit alongside—not instead of—changes to the work itself.

A 2023 overview of systematic reviews found moderate to strong evidence that organisational-level interventions can improve working conditions and employee outcomes, particularly interventions involving working-time arrangements, influence over work tasks and organisation, healthcare work design changes, and improvements to the psychosocial work environment. It also found positive evidence for participative approaches, better communication and support, and increased employee involvement or control.

A practical control table for leaders

What leaders should do now

A defensible psychosocial risk program should look like a WHS risk-management system, not a benefits calendar.

First, identify the hazards. Use worker consultation, health and safety representatives, surveys, exit data, grievance themes, workers compensation data, absenteeism, overtime, turnover, incident reports and team discussions. The focus should be on the design and management of work, not simply on whether individual workers are coping.

Second, assess which hazards are causing or likely to cause harm. Look for patterns: one team with chronic overtime, one manager with repeated complaints, one role with constant exposure to aggression, one department with unclear priorities, or one change program driving burnout.

Third, apply higher-order controls. Eliminate unnecessary hazards where possible. Where elimination is not reasonably practicable, minimise risk by redesigning work, improving staffing, changing systems, clarifying roles, adjusting rosters, improving supervision, preventing harmful behaviours and managing traumatic exposure.

Fourth, consult workers before choosing controls. Workers often know exactly where the job is breaking down. Consultation also helps prevent leadership from choosing attractive but ineffective solutions, such as a wellbeing day when the real problem is impossible workload.

Fifth, make managers accountable for conditions, not just engagement scores. Leaders should be measured on workload sustainability, turnover, overtime, grievance trends, injury trends, team climate, response times to complaints and whether agreed controls were implemented.

Sixth, evaluate whether controls work. A psychosocial risk control should have an owner, a due date, a review point and evidence of effectiveness. For example: Did overtime reduce? Did aggressive incidents fall? Did claims reduce? Did role clarity improve? Did workers report better support? Did the control remain in place during peak demand?

Finally, keep wellbeing initiatives in their proper place. EAPs, counselling, yoga, social events, mental health literacy and wellbeing activities can be useful supports. They can help with recovery, connection and early help-seeking. But they should sit behind the main WHS question: what are we changing in the work so people are not harmed in the first place?

Bottom line

Wellness programs are a “nice to have” when they build connection, recovery and morale. They become a problem when leaders treat them as proof that psychosocial hazards are being managed.

For legal defensibility and genuine prevention, leaders need to show that they have identified the psychosocial hazards in their workplace, consulted workers, applied the hierarchy of controls, changed the systems of work that create harm, monitored whether the controls worked, and adjusted when the evidence showed they did not.

The best test is simple: does this intervention change the source of the risk, or does it merely help people tolerate the risk for longer?