Safety Incentives: 4 Distortions That Reward Silence Instead of Safer Behavior

A supervisor receives a team award because the shift has reported no incidents for six months. During the same period, two operators quietly bypassed a damaged guard, a contractor declined to document a recurring lifting problem, and a near miss was closed after a reminder was sent. The dashboard looks calm. The work is not.

This is the central problem with safety incentives. A reward designed to reinforce care can teach people that the safest career move is to keep bad news invisible. As Andreza Araujo has observed across more than 250 cultural transformation projects, people read the consequence attached to a behavior faster than they read the value printed on a poster.

Safety incentives are useful only when they reinforce the decisions that reduce exposure. When they reward a perfect record, rapid closure, or visible participation without checking the quality of the control, they create a performance signal that conflicts with safer work. The four distortions below show where that conflict begins and how plant leaders can correct it.

Why a perfect safety record can become a weak signal

A low incident count is relevant, but it does not explain why the count is low. Exposure may have fallen, controls may have improved, work may have moved elsewhere, or people may have stopped reporting. The number alone cannot distinguish those conditions.

James Reason's work on organizational accidents helps explain the danger. Harm emerges when several defenses contain weaknesses that line up. A reward based only on the final outcome pays attention to the last visible event while ignoring the conditions that make that event more or less likely.

That is why the first question for a leader should not be whether the team qualified for the bonus. It should be what evidence shows that the exposure changed. If the answer is only a low incident total, the incentive is measuring silence and luck together.

Distortion 1: zero incidents becomes the team objective

A zero-incident target sounds ambitious because it expresses a serious concern for people. The distortion begins when the target becomes a team competition and every report threatens the group reward. In that setting, the incentive does not ask whether work is becoming safer. It asks whether the record can remain clean.

Supervisors then develop informal filters. A first-aid case becomes a private conversation. A near miss becomes a coaching note with no record. A recurring equipment defect becomes a maintenance request that never enters the safety system. None of these choices requires bad intent. The program has simply made reporting expensive.

Andreza's book Safety Culture: From Theory to Practice treats culture as something visible in repeated decisions, not as a statement of belief. If the repeated decision is to protect the score before protecting the signal, the incentive is training the wrong habit.

A safer design keeps incident data in the review, but it does not use a perfect record as an automatic reward trigger. The review should examine reporting credibility, the quality of escalation, and whether critical controls were verified under real operating conditions.

Distortion 2: more observations are mistaken for better behavior

Observation campaigns often produce impressive numbers. A plant reports thousands of conversations, cards, or observations, and the volume is presented as proof that people are engaged. Volume can be useful, although it becomes misleading when the program rewards the count rather than the quality of the decision.

Workers learn to submit easy observations because an empty form is safer than a difficult conversation. A missing sign is recorded repeatedly, while a production pressure that changes how a task is performed remains untouched. The program creates activity without increasing the organization's ability to see exposure.

The difference becomes clearer when leaders compare a high-volume observation program with the failures that make behavioral observation decorative. A useful observation describes the work, the control at risk, the decision made, and what changed afterward. It is not merely a count of interactions.

Recognition should therefore focus on evidence quality. A supervisor who documents one recurring exposure, escalates it to the right owner, and verifies the correction has created more value than a team that submits fifty low-consequence observations with no control change.

Distortion 3: fast action closure hides weak corrective work

Action-aging dashboards create another attractive but dangerous incentive. Leaders want overdue actions to fall, so teams are praised when they close actions quickly. The metric improves, yet the underlying risk may remain because the action was closed with a memo, a toolbox talk, or a training record.

Fast closure is not the same as effective closure. A corrective action should change a condition, decision right, barrier, or verification practice that contributed to the exposure. When the closure evidence only proves that someone communicated, the organization has reduced administrative backlog rather than risk.

The training record problem is especially common. As the article on assumptions hidden by safety training completion explains, completion proves attendance more easily than competence. The same logic applies to corrective actions. A signature proves a step occurred, not that the work became safer.

Redesign the reward around verified effectiveness. Give credit when the responsible owner can show that the control works after implementation, that the affected workers can use it under normal pressure, and that the action remains effective when conditions change.

Distortion 4: individual recognition shifts attention away from system conditions

Personal recognition can reinforce care when it celebrates a responsible decision. It becomes harmful when it implies that safety depends mainly on individual alertness, courage, or compliance. A worker may be praised for catching a hand-injury hazard even though the equipment design repeatedly creates the same exposure.

This pattern can turn a structural problem into a personal success story. The employee is recognized, the event is closed, and the organization keeps the design that made the intervention necessary. Recognition has then absorbed the energy that should have funded engineering control.

Across 25+ years of executive EHS work, Andreza Araujo has connected safe behavior to the conditions in which people make decisions. During her PepsiCo South America tenure, when the accident ratio fell 50% in six months, the relevant lesson was not that workers needed more applause. The lesson was that leadership attention, operating discipline, and field decisions had to reinforce one another.

Individual recognition should name the decision and the system response. A strong example is, “Maria stopped the lift, escalated the damaged sling, and the area owner replaced the control before work resumed.” The organization should then show what it changed so that the next worker does not need exceptional vigilance to compensate for the same weakness.

What responsible safety recognition looks like in the field

A useful recognition system rewards behaviors that increase the quality of information and the strength of controls. It does not require every person to produce the same behavior because different roles have different authority, exposure, and decision responsibilities.

For operators, that may mean reporting an uncertain condition with enough detail for action. For supervisors, it may mean pausing a task, resolving a conflict between production and control, or escalating a risk that cannot be managed locally. For managers, it may mean funding a design change and verifying that the change works across shifts.

The scripts that make safety conversations more specific are useful here because recognition should refer to an observable decision rather than a vague attitude. “You showed commitment” is difficult to learn from. “You identified the missing isolation point, stopped the handover, and brought the area owner into the decision” tells the team what good judgment looked like.

How leaders can test an incentive before scaling it

Run a short review with the people who receive the score. Ask what they believe will damage their result, what they would avoid reporting, and which actions receive praise even when the exposure remains. Their answers reveal the operating incentive more honestly than the program document.

Then compare five evidence streams over a defined period. Review reporting volume and detail, time from signal to escalation, action closure evidence, field verification results, and the number of decisions that required production pressure to be resolved. The purpose is not to create another scorecard. It is to test whether the incentive is improving the decision system.

Leaders should also sample the quiet cases. Speak with a worker whose task had no incident, no observation, and no corrective action. Ask what made the task difficult, which control was least reliable, and what the person believed would happen after reporting it. Silence is data when the organization has made visibility costly.

The supervisor curiosity routine gives frontline leaders a practical way to ask these questions without turning every interaction into an audit. The key is to connect curiosity to a decision that can be improved, not to collect another favorable response.

The leadership decision is what the program makes visible

Every incentive system declares a priority through its consequences. If people receive more recognition for a clean record than for escalating a serious concern, the program has declared that appearance outranks evidence. If a closed action matters more than a verified control, the program has declared that administrative speed outranks risk reduction.

Safety leaders can reverse that message by publishing the decisions they want repeated. Show that a credible report protected a team from pressure, that a stop-work decision received support, and that an owner was held accountable for a control that failed. When people see those decisions produce fair and visible outcomes, reporting becomes part of performance rather than a threat to it.

That is the practical distinction between a safety incentive and a safety decoration. The first changes what the organization funds, escalates, verifies, and discusses. The second creates a favorable story around numbers that may never describe the work.

What to change in the next review cycle

Remove automatic rewards tied to a perfect incident record, then replace them with a review of evidence quality and control effectiveness. Keep the incident count as context, because outcomes still matter, but do not let it decide the reward without examining exposure, reporting behavior, and response quality.

Require every recognized action to answer four questions. What risk was visible? Who had authority to act? What changed in the work? What evidence confirms that the change holds under pressure? These questions keep recognition connected to the system rather than to personal image.

Finally, ask leaders to model the behavior publicly. When a manager thanks a team for reporting an inconvenient condition, funds the correction, and returns to verify it, the organization receives a stronger lesson than any poster or bonus can provide.

Conclusion: reward the decision that makes risk harder to repeat

Safety incentives fail when they make the absence of bad news more valuable than the presence of good evidence. The correction is not to abandon recognition. It is to reward responsible reporting, proportionate escalation, effective controls, and decisions that reduce the need for individual heroism.

Andreza Araujo's work, including the experience documented through 250+ cultural transformation projects, points to a practical standard. A safety program is credible when the behavior it praises matches the decision it needs from people under pressure. If your organization wants to redesign its safety incentives around field evidence and control ownership, Andreza Araujo's team works with leaders on that transition.