Safety Incentive Programs: 7 Distortions Leaders Must Fix

Safety incentive programs look harmless when the award is small, the banner is cheerful, and the dashboard shows another month without a recordable injury. The problem is that incentives teach faster than speeches. Workers notice what gets praised, what gets paid, what gets questioned, and what makes the team lose recognition.

This article is written for EHS managers, executives, and site leaders who need to reward safe work without training people to hide bad news.

Why safety incentives can distort the truth

The common mistake is treating a safety incentive as a communication tool when it is actually a measurement system with social consequences. If the reward depends on no injuries, no near misses, or no complaints, the organization has created a reason to keep weak signals quiet.

OSHA recordkeeping rules and ISO 45001:2018 both assume that organizations can see events, investigate them, and improve controls. An incentive that suppresses reports attacks that assumption from the inside because the formal system still exists, while the field learns that silence is safer than transparency.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture appears in repeated decisions under pressure. Incentives are one of those decisions. They tell people whether leaders want accurate risk information or a clean month.

The thesis is direct. A safety incentive program is acceptable only when it increases risk visibility and control quality. If it mainly rewards the absence of reported harm, it may be buying silence.

1. Stop rewarding luck as if it were control

A team can work with weak controls for weeks and still avoid injury. That is not proof of safety. It may be luck, low exposure, small sample size, or a period in which the most dangerous tasks did not occur.

When the incentive rewards the absence of injury, leaders may celebrate a result that no one actually controlled. The crew receives the message that the outcome matters more than the decisions that produced or failed to produce it.

A stronger program rewards evidence that the team controlled the work. Examples include verified isolations, high-quality pre-task risk assessments, corrected line-of-fire exposure, closed critical-control gaps, and escalated defects before the job began.

This connects with leading indicators that TRIR will never show, because the better measure is whether people acted before harm rather than whether harm happened to stay absent.

2. Test whether reporting fell after the reward started

The first distortion test is simple. Compare reporting behavior before and after the incentive began. If near misses, first-aid reports, hazard observations, or anonymous concerns fall sharply while exposure is unchanged, leaders should assume the program may be influencing disclosure.

Underreporting rarely announces itself. It appears as a cleaner dashboard, shorter descriptions, fewer uncomfortable details, and fewer repeated concerns from contractors or temporary workers. The number improves while the quality of risk intelligence deteriorates.

Andreza Araujo's Portuguese title Muito Alem do Zero, commonly explained in English as Far Beyond Zero, warns against treating zero as a complete truth. That warning applies directly to incentives because zero can become a target that people protect instead of a result that leaders interrogate.

Use underreporting signals as a required companion metric. A reward should never be approved without a parallel test for reporting suppression.

3. Reward report quality, not report volume alone

Some companies react to underreporting by rewarding the number of reports. That can create the opposite distortion. People may submit low-value observations, duplicate minor issues, or write vague notes that inflate activity without improving control.

The incentive should reward report quality. A useful report names the exposure, describes the condition, identifies the control that is missing or weak, gives enough context for action, and receives feedback from a decision owner.

Across 25+ years leading EHS in multinational environments, Andreza Araujo has observed that people keep reporting when they see action, not when they see a slogan. The reward should therefore recognize both the person who raises the signal and the leader who converts it into a field decision.

The same principle strengthens speak-up metrics. Voice is valuable when it changes decisions, controls, and learning speed.

4. Protect the person who reports bad news

An incentive program becomes culturally dangerous when one worker's report can cost the whole team a prize. Even when leaders say reporting is encouraged, the peer pressure created by the reward can punish the person who tells the truth.

The design should separate reporting from penalty. A team should not lose recognition because someone reports a near miss, discomfort, defect, injury, or control weakness. The team may lose recognition if it hides the event, retaliates, delays escalation, or fails to act on a known exposure.

James Reason's work on latent conditions helps leaders avoid the lazy conclusion that the reporter created the problem. Most reports reveal conditions that already existed. The person who reports has made the risk visible.

Link this test with bad-news responses that protect voice. The first leadership reaction after a report often determines whether the next report arrives early or not at all.

5. Include SIF exposure even when injuries are low

A low injury count can coexist with serious injury and fatality exposure. A team may have no recordables while working near mobile equipment, energized systems, suspended loads, confined spaces, chemical transfer, or work at height with weak controls.

The incentive should therefore include SIF precursor review. Did the team identify serious-risk pathways? Did it verify critical controls? Did supervisors stop or redesign work when a fatal-risk barrier was missing? Did leaders fund the correction when the issue exceeded site authority?

Without that layer, the program may reward teams that were fortunate around severe exposure while ignoring teams that created uncomfortable visibility. That is backwards. The team that reports a serious weak signal may be protecting the organization more than the team that stays quiet.

Use SIF precursor metrics before any executive recognition. Serious-risk visibility should count in favor of the team when it leads to stronger controls.

6. Measure supervisor response after the incentive

Incentives do not operate only on workers. They also shape supervisors. A supervisor who is judged on clean numbers may discourage reporting, reclassify events, delay escalation, or solve issues informally so the score stays intact.

Measure supervisor response quality. Did the supervisor thank the reporter, protect the person from peer pressure, verify the condition, escalate the issue, and close the loop? Did the supervisor explain why a report helps the team rather than harms it?

In more than 250 cultural-transformation projects supported by Andreza Araujo's work, a repeated pattern appears. The supervisor is the translator between official intent and field reality. If the supervisor treats the incentive as a number to defend, the program will drift toward silence.

This belongs beside safety conversations, because rewards change behavior only when daily leadership language supports the right behavior.

7. Put incentive distortion on the executive dashboard

Executives should not receive only the list of teams that earned recognition. They should see the distortion tests beside the results. That includes reporting quality, near-miss trend, repeated defects, SIF precursors, corrective action quality, and worker trust signals.

The executive question is not whether the incentive made the dashboard look better. The question is whether the incentive helped the organization see risk earlier and correct it faster.

During Andreza Araujo's tenure at PepsiCo South America, where the accident ratio fell 50% in six months, the relevant lesson is that improvement depended on leadership discipline and operating routines, not cosmetic scorekeeping. Incentives should support those routines rather than replace them.

Connect this review with the executive safety dashboard. If the incentive data cannot survive an underreporting challenge, it is not board-ready.

Safety incentive design matrix

What to change before the next award cycle

Before the next award cycle, rewrite the incentive criteria so that reporting cannot harm the person who tells the truth. Then add three positive criteria: one for quality reporting, one for verified control improvement, and one for leadership response speed.

Review the last ninety days of reports, near misses, stop-work decisions, and corrective actions. If the incentive winners also show weak reporting quality, repeated defects, or unresolved serious-risk exposure, the program is rewarding the wrong signal.

Ask workers one direct question during field conversations: what would make someone hesitate to report a problem this month? If the answer mentions the reward, recognition, team pressure, supervisor reaction, or fear of losing status, the incentive has become a cultural control that works in the wrong direction.

Every reward cycle that celebrates silence makes the next serious signal harder to hear.

Conclusion

Safety incentive programs should reward the behaviors that make risk visible and controllable. They should not reward a clean number that workers learn to protect.

Safety is about coming home, and that requires truth before celebration. If your organization needs to test whether its incentives are improving safety or hiding risk, Andreza Araujo and ACS Global Ventures can support a practical diagnostic at Andreza Araujo.