When the Board Asks the Wrong Question: Safety Governance and the Metrics That Mislead

A multinational mining company reported zero fatalities and an LTIFR of 0.8 at its annual board meeting. The CEO described it as "our best safety year on record." The board minutes recorded congratulations. Bonuses were paid against the safety KPI. Three months later, a worker died in a confined space entry. The investigation revealed: the confined space entry procedure had not been reviewed in four years. The atmospheric monitoring equipment had not been calibrated in eight months. The rescue plan had never been tested. The permit-to-work system had a gap that allowed entry without a second verifier. None of these failures were visible in the LTIFR. The metric measured outcomes — injuries that resulted in lost time — not the health of the management system that was supposed to prevent them. The board was governing safety by looking in the rear-view mirror and congratulating itself on the view.

THE LTIFR PARADOX Safe Work Australia formally acknowledged in 2025 that LTIFR has critical limitations: it fails to measure actual injury frequency, it correlates poorly with severity, and it systematically excludes near-catastrophic events. A near-miss that could have killed ten people has zero impact on LTIFR. A minor strain that causes one day off work increases it.

1. "What is our LTIFR?": This measures a subset of past injuries. It tells the board nothing about current risk exposure, the health of critical controls, or the probability of a serious event. A low LTIFR can coexist with catastrophic risk.

2. "Are we compliant?": Compliance measures whether the organisation meets minimum regulatory requirements. It does not measure whether the organisation is safe. Every major disaster in industrial history occurred in organisations that were compliant at the time of the event.

3. "How many incidents did we have?": Incident count without severity weighting, causation analysis, or trend interpretation is noise, not signal. Ten minor first-aid cases and one near-catastrophic process safety event are not equivalent, but they appear so in a raw count.

4. "What are our recommendations?": Boards review recommendation lists without asking whether previous recommendations were implemented, verified, and effective. The question treats recommendations as evidence of action, when they may be evidence of inaction.

5. "Is our safety culture good?": Safety culture surveys measure perception, not practice. An organisation can score highly on culture surveys while its management system has critical gaps. Culture is the residue of what people actually do, not what they say they believe.

KEY TAKEAWAY If your board's safety governance consists of reviewing a single number (LTIFR or TRIR) and asking "are we compliant?", your board is not governing safety. It is governing the appearance of safety. The distinction costs lives.

1. "What are our top 5 SIF-potential exposures and the status of their critical controls?": This forces the organisation to identify where a fatality could occur and whether the barriers are functioning. It is forward-looking, not backward-looking.

2. "What did our last three investigations reveal about our management system?": This shifts the board's attention from incident count to systemic learning. It asks whether investigations are producing change, not just reports.

3. "What is the verified implementation rate of recommendations at 90 days?": This measures whether the organisation actually changes after an investigation. A rate below 70% indicates a systemic failure in post-investigation governance.

4. "What percentage of our workforce has current competency verification?": This measures the front-line defence. If competency verification is incomplete, the organisation is relying on luck, not capability. Includes contractors.

5. "When was the last time we tested our emergency response for a realistic scenario?": An emergency response plan that has never been tested under realistic conditions is a hypothesis. The board should know when it was last proven.

THE GOVERNANCE SHIFT The fundamental shift is from outcome-based governance ("what happened?") to system-based governance ("are our defences functioning?"). Outcome metrics tell you about the past. System metrics tell you about the future.

"When the board asks 'what is our LTIFR?' and receives a low number, they believe they are governing safety. They are governing a spreadsheet. The risks that will kill next are not in the spreadsheet."

— Bruno Hounkpati