Rewarding
zero

A zero-accident target sounds like an excellent basis for rewarding occupational safety. When no accidents occur, employees receive a reward and the company gets a safer workplace. Everyone appears to win.

Yet a serious problem can be hidden inside a safety incentive. The system may encourage people to prevent accidents, but it may also encourage them not to report those accidents.

About fifteen years ago, I designed an incentive model that initially looked exceptionally successful. The number of reported accidents collapsed, the largest factory came close to zero, and both management and employees were satisfied.

Only later did I have to ask what the system was really measuring: safety or silence?

Employees wanted time as their reward

Designing the safety incentive was part of my MBA thesis. I did not want to select a reward based on management assumptions. I wanted to find out what employees actually valued.

I interviewed more than one hundred employees at different locations. I asked what kind of reward they would consider valuable if the company wanted to recognise a good safety result.

The answer was surprisingly clear. Most did not want money, gift cards or material prizes. They wanted time for their families, hobbies and life outside work.

We therefore chose additional paid days off as the reward.

The solution felt well considered. It reflected employees’ own wishes, it was easy to understand and it had genuine value for them. Because the reward was not money, I also believed I had avoided some of the problems associated with financial incentives.

When the system was introduced, the results improved quickly. Reported accidents fell to a fraction of their previous level. The change was especially pronounced at the company’s largest factory, where safety development had long been slow. During the trial year, the factory came close to zero reported accidents.

According to the statistics, this was a breakthrough. Management was pleased, employees received the time off they wanted, and I believed I had designed an effective safety incentive.

If the evaluation had ended there, the system would have looked like an indisputable success.

A minor injury revealed the problem

A safety metric may look objective, but its meaning always depends on how the underlying information was created. Accident statistics include only the events that reach the reporting system.

The harmful effects of the incentive did not appear immediately. I visited the sites and spoke with employees regularly, but at first nobody criticised the system. People seemed genuinely satisfied with it.

Gradually, I began to hear comments and observations that did not quite fit the picture presented by the statistics. Eventually, one conversation changed my understanding of the entire system.

An employee had suffered a very minor injury: a small scratch that would not have caused incapacity or absence. The injury itself was trivial. What followed was not.

When the employee considered reporting the injury, coworkers tried to silence him. If the accident were recorded, the entire group’s reward could be at risk. One person’s report could mean that everyone else lost their additional paid days off.

Persuasion eventually became intimidation. The employee was made to understand that reporting the injury could lead to trouble outside working hours.

The system had been intended to improve safety, but it had also turned injury reporting into a social risk. Reporting was no longer only the injured employee’s concern, because it could take the reward away from the whole group.

A zero-accident target can distort reporting

People adapt their behaviour to what an organisation measures and rewards. This does not necessarily mean deliberate dishonesty. It is a natural response to the incentives created by the system.

If a reward is tied to the number of reported accidents, a good result can be achieved in two ways. Accidents can be prevented, or they can be left unreported.

The first improves safety. The second improves the statistic.

Research has found that employees may fail to report injuries and occupational illnesses because they fear consequences, experience social pressure or worry about how the information will be used. In my case, the employees’ accounts showed just how concrete that pressure could become.

The more important the zero result becomes, the harder it may be to deliver bad news. Yet an organisation needs information about minor injuries, near misses and dangerous conditions so that it can act before a more serious event occurs.

Spreadsheets and monthly reports did not reveal the problem in my system. The information came from employees who trusted that they would be heard even when their message challenged a solution management considered successful. Without that trust, the harmful effects might have remained completely invisible.

The system ended, but the improvements remained

After approximately one year, I recommended ending the incentive system.

The decision was not easy. Reported accidents had fallen substantially and the largest factory had come close to zero. However, the pressure not to report was too serious a side effect to ignore. If a safety system begins to prevent an organisation from receiving safety information, the system itself becomes part of the problem.

After the incentive was removed, the number of reported accidents did not return to its previous level. Other changes were also visible at the workplaces without having to infer them from statistics.

Housekeeping and order had improved. Personal protective equipment was used more consistently, working practices had developed, and dangerous shortcuts were less common. Employees paid more attention to hazards and made more suggestions for improving safety.

Not all of the change had therefore been a reporting distortion. Genuine safety improvement had also taken place.

It is possible that the reward helped initiate the change. As safer practices gradually became normal, maintaining them no longer depended on extra days off. People had adopted a better way of working and did not want to return to the old one.

The system cannot honestly be described as either a complete success or a complete failure. It appeared to improve both the working environment and safe behaviour, but it also influenced accident reporting.

This is precisely what makes safety leadership difficult. The same intervention can have both intended and harmful consequences.

What should we reward in safety?

A zero-accident target can express an organisation’s ambition, but it is problematic as the basis for a reward. An employee or team does not control every factor that contributes to an accident. People can, however, directly influence whether an event enters the reporting system.

The problem is not rewarding people as such. The problem is what we reward and on what basis.

Replacing accident numbers with another simple numerical measure does not solve the problem. If a reward is tied to the number of safety observations, for example, an organisation may receive a large quantity of observations without removing its most important hazards.

Rewards should focus attention on action and learning: are significant hazards identified, are they solved in practice, are agreed improvements implemented, and can employees participate in developing safer work?

A good incentive should make raising problems attractive and safe. It should never give anyone a reason to hide injuries, dangerous situations or shortcomings.

Before introducing a safety incentive, leaders should therefore ask one difficult question:

What behaviour does this system actually encourage?

The answer is not found in the system’s stated objective or presentation material. It is found in what people can gain or lose, and in the choices the reward makes attractive to them.

A zero-accident reward can make safety statistics look excellent. That does not yet prove that the workplace is safe.