A May 12, 2012, report by the Government Accounting Office “Better OSHA Guidance Needed on Safety Incentive Programs” raises some concerns about incentive programs and the way that many companies are using them. OSHA recommends behavior-based incentives and can prohibit rate-based incentives when they are part of an OSHA cooperative program, but otherwise has no regulatory authority over how incentive programs are designed and implemented. On the other hand, there is a lot we know about incentive programs that demonstrates why companies should avoid rate-based metrics regardless of the regulatory requirements.

The problem with injury-rate metrics is that they are based on outcomes and then they award tangible incentives. Both of these have been shown time and time again to be counterproductive for safety — and yet they are among the most commonly used by companies in a variety of industries. My suspicion is that there is a simple explanation — outcomes are the easiest to measure and tangible incentives are the easiest to give out. Let’s review both of these principles to make sure we all know the best practices for safety incentive programs. These are not the only two factors that need to be considered when designing a program, but according to the GAO report, they are a frequent source of mistakes.                   

What should incentive program rewards be based on?

Obviously, incentive programs need to be based on something. How should we decide who gets a reward? Outcomes are easy to measure, transparent to employees, and cheap. We had 40 injuries last year. This year we had 30. Sounds like an improvement to me — hand out the rewards!

There are three reasons why this is not nearly so simple. First, the context might have changed. Maybe we laid off half the workforce and there are fewer workers in the baseline. Maybe the equipment is a lot safer this year and a decrease from 40 to 30 is actually less than we should have expected. These challenges can be addressed by using rates rather than absolute values. The rates can be adjusted to account for changes in the workforce, the facility, throughput rates, or other known factors.

The second reason is that injuries are often the result of stochastic effects (the unknown factors). A worker can do the same activity a dozen times using an unsafe method and never get hurt. Another worker can do the same activity using the safe method just once and get hurt. Why should the first employee get the reward but not the second? To be effective in creating long-term sustainable change, the incentive program needs to create a logical link between safe behavior and success.

And then there is an even more insidious problem with outcome-based measures. When an employee can get a reward for avoiding injuries, there is an equal incentive for hiding an injury as there is for preventing it. When employees hide a real injury, their productivity goes down and the injury gets more expensive to treat when it finally is reported. If they conceal it until they just can’t work anymore, the company now has additional costs for replacing the worker and training a replacement. There is also a hit to morale when a co-worker goes on disability.

What kinds of rewards should be granted?

It makes sense to think that the rewards employees would appreciate the most are the ones with a real tangible value. Cash, gift cards and swag immediately come to mind. These are also easy, transparent and simple.

The problem with tangible rewards is the effect they have on the relationship between the employee and safety. When people choose a behavior to earn a tangible reward such as cash, it becomes an economic transaction. Their focus shifts to earning the cash rather than a desire to be safe.

Instead, organizations should try to maintain safety as its own value. We want employees to be safe because it is important — for themselves, for their team and for the company. When safety is kept within this social frame, even employees who don’t receive the reward still want to be safe just for safety’s sake. They still want to be a good person, a good teammate and a good employee. Safety becomes part of their self-identity.

A simple example of this is illustrated by classical studies of cognitive dissonance. In a typical case, participants were recruited to do a really boring activity. Half of the recruits were given a cash payment for their participation. The other half was prominently thanked for their contribution to scientific research. Guess which half rated their participation as a worthwhile use of their time and were more likely to do it again? It was not the half who got the cash. It was the half who felt good about themselves for participating.

So now what?

So what are the alternatives? What is there besides safety outcomes to base rewards on? What other kinds of rewards are there?

The option recommended by OSHA and much of the research on safety behavior is to use employee behavior. If an employee is behaving safely, he or she should earn the reward regardless of the result; and the employee who is behaving unsafely, risking his or her own safety and exerting a negative influence on others, should not.

Then the reward should be something intangible that maintains the social frame of the employee-safety relationship. This could include anything that is linked to the intrinsic value of safety. Public recognition of the good safety record is a good place to start. Assign the safe worker to a position as safety trainer or safety mentor. Recognize them publicly to their co-workers or externally to their family or to the general public in a local paper. Make them proud to be the safe worker.

More effort, better results

Why are so many organizations making the mistakes highlighted in the OSHA report? It is too easy to take the easy way out. But if the easy way isn’t effective at improving safety, then even the smallest investment of time and money is wasted. Organizations are better served to put a little more thought, a little more effort into designing a safety incentive program that will work. The large increase in benefits is worth a small increase in investment. And once the organization develops some expertise in designing effective programs, the next one will not cost more at all.