Safety breaches

Safety breaches are valid-reason cases with an extra dimension: the Commission weighs not just what the employee did, but what could have happened, how the rule has actually been enforced, and what the employer owes everyone else on site.

What you’ll be able to do

• Explain why a safety-critical industry or role raises the bar for a valid reason, and write that into a termination letter.
• Separate “this caused real harm” from “this could have caused harm” without using either to excuse a breach or over-punish a near miss.
• Spot the inconsistent-enforcement and seniority arguments that most often decide safety-breach cases at the Fair Work Commission.

Safety sits inside s.387(a), not beside it

Section 387(a) doesn’t treat safety as a separate factor — it is folded into the valid reason test itself, which asks whether there was a valid reason for the dismissal related to the person’s capacity or conduct, “including its effect on the safety and welfare of other employees” [2020] FWC 250 at [61]. The general valid reason test is covered in Module 5; this module is about what changes when the conduct is a safety breach.

A useful Full Bench statement, quoted in Plesiotis v Huhtamaki, is that the “fair go all round” standard behind s.387 needs to weigh the employer’s own safety obligations alongside the employee’s conduct, “especially in inherently dangerous workplaces” [2020] FWC 120 at [131]. That authority (BHP Coal Pty Ltd t/a BMA v Schmidt) is not independently located in this corpus pass — check the quote against the original before relying on it.

The safety-critical nature of the role raises the bar

The more dangerous the environment, the less room there is for an employer to tolerate a breach — and the more weight a long, clean record loses against it.

• In a glass factory with “significant inherent risk from a safety perspective” [2020] FWC 250 at [11], a driver’s 30-year service and age weighed towards harshness, but the Commission found this was outweighed by “the criticality of safety in a high-risk work environment” [2020] FWC 250 at [176].
• In an electricity network crew where a customer received an electric shock, the Commission put it directly: “in an industry where the risks of injury and death are so high” the employer needs confidence that policies are followed every day, not most days [2022] FWC 1522 at [235].
• Even where the breach is described as a “near miss”, the starting point in mining is that “the mining industry is an inherently dangerous industry” [2021] FWC 838 at [148] — which is why the Commission still found a valid reason there despite no injury.

For advisers: name the hazard class in the termination letter (confined space, mobile plant, electrical isolation, patient handling, working at heights) rather than leaving “safety breach” unqualified. The seriousness of the reason stands or falls on that specific risk, not a general reference to “safety”.

Actual harm versus a near miss: both cut the wrong way if you let them

Two errors recur. The first is treating “no one was hurt” as a reason to go easy. The second is treating a near miss as automatically as serious as a breach that actually caused harm, with no further inquiry.

On the first: in the Viridian Glass driveway incident, the Commission expressly refused to grade the risk by what actually happened rather than what could have happened, because scoring by actual outcome “would then result in the downplaying of the significance or reporting of near misses” [2020] FWC 250 at [120]. A near miss is assessed on its potential consequence, not on the fact that this time nothing happened to go wrong.

On the second: in a warehouse tugger collision with no injury and no property damage, the Commission found there was no valid reason for dismissal at all [2020] FWC 3924 at [197], in part because the incident genuinely “did not result in injury to either person” and was not a high-potential incident [2020] FWC 3924 at [196]. Severity still matters — it’s the actual-versus-potential framing that doesn’t excuse a breach, not the underlying facts of how dangerous this particular incident really was.

The practical test: ask what could have happened on a worse version of these facts (closing speed, load weight, who was standing where), not only what did happen. Then separately ask whether the facts, properly assessed, were actually serious enough to justify dismissal — a low-potential incident doesn’t become high-potential just by calling it a “safety breach”.

Consistency of enforcement

Safety rules that have been enforced patchily in the past don’t lose their force — but the employer carries the inconsistency, and the Commission will weigh it.

• An employer is entitled to tighten enforcement of a rule it had let slide, including after a near miss prompts a crackdown [2020] FWC 120 at [130]. But that doesn’t make the earlier inconsistency irrelevant — it’s “certainly relevant to whether the Applicant has been treated in a manner consistent with other employees” [2020] FWC 120 at [129], and in that case it was found to weigh, on its own, in favour of unfairness [2020] FWC 120 at [156].
• Comparators have to be genuinely comparable. A coal mine employee dismissed over a “near miss” succeeded partly because two colleagues involved in an actual collision — a more serious incident — were treated more leniently: the differential treatment “renders the termination of the Applicant unreasonable” [2021] FWC 838 at [152].
• But not every different outcome is inconsistency. A warehouse operator’s dismissal survived a consistency challenge because the comparator cases involved different incidents, different histories and no “cookie cutter” policy was being applied — the employer “properly, considers each matter separately, investigates the matter and takes action considered appropriate in the circumstances” [2020] FWC 3924 at [211]. Even the one directly comparable co-worker from the same incident wasn’t a useful comparator, because he had under a year’s service and no prior record against the Applicant’s “close to 30 years” and recent history [2020] FWC 3924 at [213].
• Seniority can itself explain a harsher outcome rather than undermine it. Where several employees were investigated over the same incident but only the most senior was dismissed, the Commission accepted that being “ultimately in charge” was the aggravating factor that justified the difference, not evidence of inconsistency [2022] FWC 1522 at [231].

For advisers: before running (or resisting) a consistency argument, get the comparator’s role, seniority, history and the nature of their incident on the table. “Someone else wasn’t sacked for a safety breach” is not an argument until you know whether it’s the same rule, the same risk and a genuinely comparable employee.

Does experience cut for the employee, or against them?

Both, depending on what it’s used to prove.

• Long service is a mitigating factor under s.387(h) generally, but in safety cases the Commission has turned it around: the long period of employment “needs to be balanced against the … reasonable expectation that such a long serving employee would be in a position to understand and abide by the rules” [2020] FWC 120 at [162].
• The same point was made more bluntly elsewhere: “long employment in the one area can lead to complacency”, and that is never an excuse for a safety breach [2020] FWC 3924 at [243].
• Honesty under questioning is credited, but it is not a free pass. A crew leader’s Show Cause response was found to be “brave and honest” [2022] FWC 1522 at [232] — the Commission weighed that genuinely in his favour, and upheld the dismissal anyway because his seniority made the underlying conduct more serious, not less.

A related risk-management point advisers should flag on its own footing: harsh, inconsistent discipline for an honestly self-reported near miss risks teaching a workforce to stop reporting near misses at all — which is the opposite of what a safety system needs. This corpus pass did not turn up an FWC decision squarely deciding a case on that chilling-effect argument, so don’t present it to a client as settled Commission doctrine. It is, however, the same underlying reasoning the Commission used in rejecting an “actual outcome only” approach to risk-rating [2020] FWC 250 at [121] — and it is a legitimate reason, independent of case law, to design discipline around the breach and the role rather than around who happened to be unlucky enough to get caught, or honest enough to self-report.

What the Commission looks for

Across these decisions, the pattern is consistent:

• The specific rule breached, and whether it is genuinely safety-related (not merely administrative dressed up as “safety”).
• What could have happened on a worse version of the same facts — not only what actually happened.
• How the rule has actually been enforced across the workforce, including against comparable employees in comparable incidents.
• Whether the employee’s role, seniority or experience made the breach more serious (greater responsibility, greater expected awareness of risk) or whether it’s being used unfairly as a stand-in for “should have known better” without evidence.
• The employee’s own response: candour and acceptance of responsibility are weighed, but they sit alongside the conduct, not instead of it.

How safety-breach dismissals fail

Adviser checklist

Practice scenario: the hoist that was “always slow”

Decisions cited

Check each paragraph reference against the published decision on the FWC website before relying on it. BHP Coal Pty Ltd t/a BMA v Schmidt is cited as quoted in a later decision and still needs checking in the original.