The people who leave tell you what the people who stay won't. Most organisations record exit interviews, file them, and never aggregate them. That is a documented hazard signal sitting in a drawer.
The pattern
Under the SafeWork NSW Code of Practice: Managing Psychosocial Hazards at Work, Step One of the risk management process is to identify psychosocial hazards. The Code lists the records a PCBU should review to do this. Among them: "absenteeism, turnover, exit interviews, sick leave data and workers' compensation claims."
That language is not unique to NSW. The Safe Work Australia Model Code of Practice, which has been adopted in most Australian jurisdictions, names the same sources. The point is consistent across the regulatory framework. Exit interviews are an input to hazard identification. They are not a nice-to-have. They are part of the evidence base a regulator expects PCBUs to be reviewing for trends.
The problem is who holds the data. Exit interviews sit with HR or People and Culture in most organisations. They are read individually, sometimes summarised quarterly, and almost never reviewed as evidence of psychosocial hazards. The WHS function, which carries the legal duty to identify and control these hazards, often never sees them.
The Code is explicit that not all psychosocial hazards will show up in incident reports. Workers leaving are sometimes the only people willing to name what was wrong. Filing those conversations without aggregating them is not hazard identification. It is record-keeping.
What this looks like in practice
A worker resigns. In their exit interview, they describe sustained overtime, a manager who publicly criticises team members, and the fact that two colleagues have already left for the same reasons. HR files the form. Six months later, a workers compensation claim is lodged for a psychological injury arising from the same team.
In the investigation that follows, the exit interview is in the file. The information was captured. It was held by the organisation. It identified a hazard the Code says PCBUs should be identifying. Nothing was done with it.
What to do about it
Three practical changes.
First, aggregate. Exit interview data needs to be reviewed as a dataset, not as individual conversations. Patterns across departments, managers, and reasons for leaving are what reveal hazards. A single exit interview is anecdote. Twelve from the same team in eighteen months is a control failure.
Second, route it. The WHS function needs visibility of exit interview themes, even if HR continues to own the conversations. The data does not need to leave the People and Culture system. The patterns do.
Third, close the loop. When exit interviews identify a hazard, the response belongs in the same risk assessment process every other hazard goes through. Document what was identified, what control was considered, what was implemented, and what changed. That is the evidence a regulator, a board, or a claims investigator will look for.
The workforce is already telling you where the hazards are. The Code already says you should be listening. The question is whether your compliance system can hear them.
Disclaimer: This article provides general information on psychosocial compliance in Australian workplaces. It does not constitute legal advice. Organisations should consult qualified professionals for advice specific to their circumstances. Information cited is sourced from Safe Work Australia and relevant state regulators as of the date of publication.
