At a recent behavior-based safety conference, I heard speakers say that they did not believe in incentives or gimmicks such as mugs, tee-shirts and the like as rewards for safe working behavior. They warned of the danger of "dangling a carrot" in front of people as a means of getting them to act safety. The implication was that safe behavior generated this way was inferior to behavior generated by other means. The words and phrases, "gimmick" and "dangling a carrot," are emotionally laden and tend to convey a superficial approach and even sinister motive behind such attempts to get people to work safely. With all due respect to the experience of those who expressed such beliefs, the issue whether incentives, gimmicks, and carrots have a place in developing an effective safety culture is not a matter of anyone’s belief, it is an empirical matter. Incentives either work or they don’t. So called "gimmicks" such as mugs, belt buckles, caps and tee-shirts either work or they don’t. There is 80 years of research on the conditions under which they work and do not work. When this body of research is ignored, it should not surprise anyone that many of the ways in which incentives are applied don’t work.
In order to understand when incentives are and are not effective, some basic elements of behavior change need to be understood. The first is positive reinforcement. As I have stated elsewhere, positive reinforcement is the most important element affecting behavior and the weakest link in behavior-based safety systems. If you look at the bulk of seminar presentations and magazine articles describing behavior-based safety implementations you find that positive reinforcement, feedback, recognition and reward (see inset) are all used interchangeably as though they refer to the same process. They don’t. In addition, a common perception in the safety area is that public praise, a pat-on-the back, telling someone that they are doing a good job, and "positive feedback" are positive reinforcers. They are not. With this poor level of understanding of the most important concept in behavior-based safety it is no wonder that there are problems and confusion about what to do and not to do.
What then is positive reinforcement? There are several important characteristics. The first is that positive reinforcement is defined as any consequence that follows a behavior and increases the frequency of that behavior in the future. While praise, a pat-on-the-back, etc. may be positive reinforcers, they also may not be for a given person in a particular situation. It is not unusual that people who have been praised in public, do things to avoid such recognition in the future because the event was embarrassing to them. One of the most common errors that people make in attempting to positively reinforce safe working behavior is to choose things that you say, do or give to everybody. While saying, "Thank you for picking that package up with your legs rather than your back," may be reinforcing to someone one time, it will rarely be reinforcing to anyone if said frequently. Also, it is not a reinforcer to everybody when said only one time. Even the most commonly used attempt to positively reinforce, the "pat-on-the-back," is not reinforcing to everybody. Everyone has a different set of reinforcers. It is one of the things that makes us different. The things that people will spend their discretionary time and money to do are endless and no two people are reinforced by exactly the same things. Therefore, you can see the error of buying coffee mugs for everybody who meets some safety goal. While there are some things that practically everyone likes, you can find individuals who will not be reinforced by them. In order to develop the most effective relationship with another person, you must first know that person’s reinforcers. The bottom line is that no matter what you use, if the person’s behavior does not increase, it was not a positive reinforcer to that person.
The second characteristic is that positive reinforcement is immediate. That means that when positive reinforcement is delivered while a person is working safely, the behaviors occurring at that time are the ones that are reinforced. It is obvious that reinforcing safe-working behavior when it is occurring is the most efficient and effective way to reinforce. This has many implications for a safety culture. It means first and foremost that, behavior-based safety is about what happens day to day. Peers are certainly in the best position to deliver positive reinforcement since they can deliver it when the person has done, or is doing, something in a safe manner.
The third is that one positive reinforcement does not make a habit. Safety is mostly about habits. Depending on the complexity of a particular safety habit, it may require hundreds of occasions of positive reinforcement to attain habit status. This has obvious implications for whether a reward, which usually occurs very infrequently, can significantly effect a safety habit.
When you understand the previously stated facts about reinforcement, it is relatively easy to see the conditions under which incentives work and do not work. Incentives can be reinforcers, but several things tend to reduce their effectiveness :
They are practically always delayed. They are rarely given when safe-working behaviors are occurring. The behavior they are intended to strengthen may have occurred months before and may be no longer occurring. In some cases because of the delay, the behavior occurring at the time of the delivery of the incentive may not be something that you would want to encourage.
The link between the behaviors and the incentive is usually weak. Most safety incentives are dispensed to everyone who participated in the effort, regardless of their rate of safe behavior. Some people work actively to do everything safely while others do not, and everyone gets the incentive if some common goal was reached. It is common that people who are habitually unsafe get the incentive simply because they were on the team and not because of their behavior. This, of course, makes the incentive less valuable to those who worked hard to get it.
Not all incentives are desirable (reinforcers) to the performers. Just because you put a company logo on a cap, jacket or key chain does not necessarily make them valuable.
Here are some tips on how to make incentives effective:
When these things are done, incentives can have motivational value for safe behavior. When any of these things are omitted, or done poorly, you can expect negative reactions and poor safety results. As a part of a system of reinforcement, incentives can add energy and excitement to a safety culture. When used without regard to reinforcement, they add little and can actually be counterproductive.
Recognition - usually some symbolic way of showing appreciation for some accomplishment such as plaques, trophies, letters of commendation, etc. Since recognition is delayed and infrequent, it is only effective as a part of a reinforcement system.
Incentive - a variety of tangible items, ranging from money to merchandise designed to effect or influence the behavior of those receiving them.
"Gimmick" - a derisive term describing incentives, but usually referring to small or inexpensive merchandise such as caps, tee shirts, coffee mugs, etc. The term is usually used by those who believe that tangible rewards do not produce genuine change, but produce behavior designed only to get the gimmick. They believe, incorrectly, that when the gimmicks stop, the behavior will also stop. If this happens the wrong behavior was reinforced.
"Dangling a carrot" – this is also a derisive term used by those opposed to the use of incentives. The term describes an old motivational device used by people who worked horses, wherein a carrot was tied to a stick that stuck out over the horse’s head. To the extent that it worked, it was only temporary because whether the horse worked hard or hardly at all the carrot was no closer to being eaten. As used with people it is described as tempting someone with a reward. It has the connotation of an unsavory practice, as in getting them to do something that they would not otherwise do, as against their principles.
Feedback - information about behavior that will allow the performer to change that behavior. Feedback can be an antecedent or a consequence. If it is used prior to performance, it is an antecedent. If it is used during or immediately after a behavior, it is a consequence. If social reinforcement is consistently paired with improved behavior as represented on a chart or graph, knowledge of the results may take on reinforcing effects. On the other hand, if the feedback is paired with punishment, it may become a negative reinforcer. If changes on the graph are not followed by either reinforcement or punishment, it will have no effect on behavior.
Reward - a tangible item, usually money or exchangeable for money, that is intended to influence behavior in a particular direction. Since most rewards are not tailored to the person receiving them, it is not surprising that they do not motivate everyone. Since most rewards are for results, without regard to the behavior used to obtain them, it should not surprise anyone that people will try to get rewards with behaviors other than those desired. The fact that people will lie, cheat and steal to get some rewards is not the fault of the rewards but the fault of the one who designed the system that allows those behaviors to be rewarded. Since rewards are not always delayed, it is possible that the behaviors that triggered the reward may not be occurring when the reward is received, which inadvertently reinforces the wrong behavior.