Safety Leadership: 6 Blind Spots That Keep Decision Rights Away From the Point of Risk

F1 critical diagnostic for EHS managers, plant leaders, and frontline supervisors

A supervisor can identify a changing hazard, stop the task, and still lack the authority needed to make the next safe decision. That gap is easy to miss because the organization may have a policy, a dashboard, and a leadership walk that all appear to support intervention.

This article examines six blind spots that move decision rights away from the point of risk. The central argument is practical: safety leadership is not proven by being visible near the work, but by making the right decision possible when conditions no longer match the plan.

Why safety leadership is a decision-rights problem

Leadership is often described through presence, communication, and personal example. Those behaviors matter, although they do not answer the question that determines whether a control survives pressure. Who can change the work when the approved method no longer fits the conditions?

ISO 45001:2018 places leadership, consultation, operational planning, and improvement inside one management system. A useful reading of that structure is that responsibility cannot end with a signed procedure. The organization must also make escalation workable for the person who encounters the exposure.

Andreza Araújo argues in Safety Culture: From Theory to Practice that the frontline leader carries, translates, and sets the tone for culture. A supervisor who is accountable for the result but excluded from the decision is carrying responsibility without control.

The same distinction appears in Make The Difference: Be a Leader in Health & Safety, where leadership is treated as action in the work rather than a message delivered above it. The test is whether the policy changes what happens during a disagreement, delay, or unexpected condition.

1. Presence is mistaken for authority

A leader may spend time on the floor, ask workers what they need, and still leave authority in the office. Presence becomes symbolic when the person closest to the hazard must wait for distant approval before changing a task, isolating equipment, or rejecting a weak control.

The blind spot appears when leadership activity is measured by visits rather than decisions. A calendar full of field walks can coexist with slow escalation, because being seen near the work does not show whether the leader removed a barrier that blocked safe execution.

Ask a supervisor to describe the last time a field observation changed the work. The answer should include the condition, the decision, the owner, and the evidence that the change held. If the answer is only that the issue was reported, leadership was present but not yet effective.

Leaders can correct this by publishing a small decision-rights map for recurring high-risk work. It should name which decisions belong to the supervisor, which require technical review, and which require operational leadership, because ambiguity is itself a delay mechanism.

2. Stop-work authority has no response path

Many organizations tell people that they may stop unsafe work. Fewer explain what happens in the next ten minutes. Without a response path, the statement sounds strong in training and feels dangerous in practice.

A credible stop must lead to a defined sequence. The work pauses, the immediate exposure is stabilized, the responsible decision maker is contacted, the condition is reviewed, and the team receives a clear decision about restart or redesign. Each step needs an owner.

When that sequence is missing, workers face two forms of uncertainty. They do not know whether the leader will support the pause, and they do not know how long the pause will last. Production pressure then turns the silence after the stop into an argument for continuing.

Use the boundaries for stop-work authority as a starting point, then test one live escalation with a supervisor. The purpose is not to create another form. It is to verify that the organization can respond before the person who raised the concern is asked to absorb the cost.

3. Informal escalation hides the real operating model

Written roles may say that any worker can raise a concern, while the operating model quietly depends on personal relationships. A supervisor who knows which manager will answer at night has more practical authority than a supervisor with the same formal role but no reliable contact.

This difference matters because informal access is fragile. It changes with shifts, vacations, contractors, and management changes. It also rewards people who already know how to navigate the organization while leaving newer workers exposed to delay.

Leaders should compare the formal escalation route with the route people actually use. Ask who receives the first call, who can authorize a temporary control, and who can release resources without waiting for a weekly meeting. The gaps reveal where the real decision rights sit.

A short escalation log can help, provided it records decisions rather than only notifications. The risk-escalation decision log becomes useful when it shows the original condition, the response time, the decision owner, and what changed in the field.

4. A green dashboard is treated as proof of control

Dashboards simplify leadership attention, which makes them valuable and hazardous at the same time. A green result can describe completed training, closed actions, or a stable incident rate without showing whether supervisors can still make a difficult decision when a control becomes inconvenient.

The blind spot is not the use of metrics. It is the substitution of metric completion for operating evidence. If an indicator confirms that actions were closed but does not show whether the barrier was tested under normal pressure, it cannot carry the full weight of a control claim.

James Reason's work on organizational accidents helps explain why this matters. Harmful events develop through layers of conditions and failed defenses, so leadership review must examine how defenses behave in work, not only whether a record says they exist.

Leaders can add one decision-quality question to the dashboard review. Which recent red signal changed a decision, and which green signal was challenged in the field? The questions make room for evidence that does not flatter the system, which is exactly what a leadership review is supposed to do.

5. Corrective actions are assigned without decision owners

Action registers often identify a person responsible for completing a task, but completion ownership is not the same as decision ownership. A coordinator may upload a revised procedure while the operating leader remains responsible for deciding whether the new control will be funded, staffed, and verified.

This distinction becomes visible after a serious near miss. The report may contain sensible actions, yet nobody can explain who had the authority to change the schedule, redesign the task, or reject the production plan that created the exposure.

Andreza Araújo's The Illusion of Compliance, translated from A Ilusão da Conformidade, separates declared conformity from protection that remains effective under pressure. An action is not strong because it is assigned. It is strong when the decision that controls exposure has changed.

For each critical action, name both the completion owner and the decision owner. Then define the field evidence that will show the change is operating. This prevents administrative closure from being mistaken for risk reduction.

6. Leaders ask for reassurance instead of friction

Questions shape the information that reaches a leader. “Is everything under control?” invites a quick yes, especially when the team expects the conversation to end. “Where is the plan hardest to execute today?” asks for the friction that a control system needs to see.

Safety leaders ask more questions than they give answers, as Andreza Araújo states in Safety Culture Diagnosis: Learn how to do your own. The point is to help supervisors expose conditions that will defeat the plan before those conditions become normal.

Useful questions connect authority to work. What decision would the team make if the permit condition changed? Who can approve a safer alternative? What pressure would make a competent person take the shortcut? Which answer would require a leader to change the plan rather than coach the worker?

A leader who only hears reassurance receives a polished version of the operation. A leader who asks about friction receives the operating model itself, including the points where people are expected to carry risk without enough time, information, or authority.

How can leaders test decision rights in one operating week?

Select one recurring task with meaningful exposure and observe it across two shifts. Do not begin with a broad culture survey. Begin with the decisions the task requires when access, timing, equipment, or staffing differs from the written plan.

Ask the person doing the task to identify the first condition that would make the plan unreliable. Then ask who can change the plan and how that authority becomes visible to the crew. If the answer depends on a private relationship or a manager who may not be available, the decision right is not yet robust.

Record the response time, the decision path, and the changed condition. A small sample is enough to expose whether the organization has built an operating capability or only written an intention.

What leaders should change before the next pressure event

Start by choosing one high-consequence task and writing its decision-rights map in plain language. Give the frontline supervisor authority to pause the exposure, define the technical boundary for restart, and state the escalation route when the control cannot be restored locally.

Then review the evidence that reaches the leadership team. Replace broad assurances with one question about friction, one question about authority, and one question about what changed after the last warning.

Use the questions for a stronger safety operating review to examine whether leadership decisions reach the work, and compare the result with the behaviors that make safety leadership credible.

The next pressure event will reveal whether the person who sees the hazard can change the decision before the exposure becomes harm.

Conclusion: move authority closer to the hazard

Safety leadership becomes credible when accountability and authority meet at the point of risk. Leaders do not need to remove every decision from the frontline. They need to make the boundaries clear, the escalation path dependable, and the response visible after someone raises a concern.

Across 25+ years leading EHS at multinationals, Andreza Araújo has treated culture as something leaders sustain through repeated decisions, not something they announce once. When decision rights move closer to the hazard and leaders remain accountable for the system around them, safety stops depending on personal courage alone.