• An Image Slideshow
  • An Image Slideshow
  • An Image Slideshow
  • An Image Slideshow
  • An Image Slideshow
  • An Image Slideshow
  • An Image Slideshow

Psychology of Behavioral Safety

Many companies have spent a lot of time and effort improving safety, usually by addressing hardware issues and installing safety management systems that include regular (e.g. monthly) line management safety audits. Over a number of years these efforts tend to produce dramatic reductions in accident rates.

Often, however, a plateau of minor accidents remains that appears to be stubbornly resistant to all efforts to remove them. Although many of these are attributed to peoples' carelessness or poor safety attitudes, most of these are triggered by deeply ingrained unsafe behaviors. Behavioral Safety addresses these by making use of proven management techniques which almost always results in a positive step change in safety performance and safety attitudes.

 Why Focus on Unsafe Behavior?

Although difficult to control, approximately 80-95 percent of all accidents are triggered by unsafe behaviors, which tend to interact with other negative features (termed Pathogens) inherent in workflow processes or present in the working environment. Often inadvertently introduced by the implementation of strategic plans, every organization has its fair share of accident causing pathogens. These pathogens lie dormant and are relatively harmless, until such time as two or more combine and are triggered by an unsafe behavior to produce an accident.

Illustrating this, is a company that installed a new production process that entailed designing and building two new mezzanine floors in an existing plant. A project team had approved plans developed by plant based engineers. Once the construction work was complete, it was found that supporting girders had been installed five foot above the second step of a staircase on both floors, thereby introducing two pathogens into the physical environment. During commissioning of the process equipment, product blockages were frequently found to occur in the related pipe work (a third pathogen) that could only be cleared by going to the top mezzanine floor where the inspection hatch was situated. Due to increased production pressures and reduced manning resulting from a downsizing exercise the blockage required the operator to isolate the equipment at a lower production floor (another pathogen), and ascend the stairs to the mezzanine floors to clear the pipe work. At this point all these pathogens combined to trigger an accident when the operator rushed up the stairs to clear the blockage. He ran into one of the low girders, gashing his head and inflicting whiplash effects on his neck while also knocking himself unconscious. This resulted in a reportable accident, lost production and associated costs, etc.

In this true example, the potential for this type of lost-time accident will always be present until such time as the pathogens are addressed. Given that it is much more difficult to address these resident pathogens, focusing attention upon the operator’s unsafe behavior of running up the stairs is a much easier option as it is within the operator's control, whereas the pathogens are not. Because behavioral safety approaches identify and focus on particular sets of unsafe behaviors, people tend to be more aware of their potential to cause harm. In turn this gives people the mechanism by which they can control their own safety behavior and that of their colleagues.

A focus upon unsafe behaviors also provides a much better index of ongoing safety performance than accident rates for two reasons: First, accidents are the end result of a causal sequence that is usually triggered by an unsafe behavior; And second, unsafe behaviors can be measured in a meaningful way on a daily basis. Accident rates tend to be used as the primary outcome measure of safety performance simply because they signal that something is wrong within the company's safety management system. Because of the way they are calculated, they also provide a crude benchmark by which companies can compare the effectiveness of their safety management systems across industries. Unfortunately, this tends to result in management attention and resources being focused on safety only when accident rates rise dramatically. When the immediate problems appear to be resolved, management attention and resources are diverted to other pressing organizational issues until such time as the accident rate rises once again, and so on.

Consequently, rather than being proactive, those who focus almost exclusively on accident rates as a measure of safety performance tend to be reactive in their approach to safety. Conversely, a regular focus on actual safety behavior is proactive as it allows other safety-related issues in the accident causal chain to be identified and dealt with before an incident occurs. Because 'safety behavior' is the unit of measurement, a collaborative, problem-solving approach involving both management and employees is adopted to identify critical sets of safe and unsafe behaviors and used to develop 'Safety Behavior Inventories' (See Cooper, 1998). These inventories provide the basis for personnel to systematically monitor and observe their colleague’s ongoing safety behavior, on a daily basis, in an enabling atmosphere. Based on the first few weeks’ results of the peer monitoring, the workforce set their own 'safety improvement' targets. Information feedback is then provided on a weekly basis to allow the workgroups to track their progress in reaching the safety improvement targets. Companies adopting this approach are usually rewarded by fewer accidents, consistent safety management, better communications and greater involvement in team working, all of which can exert beneficial effects on production related issues and bottom line profits.

Why Do People Behave Unsafely?

People often behave unsafely because they have never been hurt before while doing their job in an unsafe way: 'I've always done the job this way' being a familiar comment. This may well be true, but the potential for an accident is never far away as illustrated by various accident triangles. Heinrich's triangle, for example, suggests that for every 330 unsafe acts, 29 will result in minor injuries and 1 in a major or lost time incident. Over an extended period of time, therefore, the lack of any injuries for those who are consistently unsafe is actually reinforcing the very behaviors that in all probability will eventually lead them to be seriously injured. The principle being illustrated here is that the consequences of behaving unsafely will nearly always determine future unsafe behavior, simply because reinforced behavior tends to be repeated.

Although it is not unusual to find the continuation of unsafe behaviors being supported by more than one reinforcer, some will exert stronger effects on peoples' behavior than others. This is particularly the case for reinforcers that are soon, certain and positive. Smokers, for example, find it hard to stop because the consequences of smoking are soon (immediate), certain (every time) and positive (a nicotine top up), whereas the negative consequences (e.g. lung cancer) are late (some years away) and uncertain (not every smoker contracts or dies from lung cancer). In exactly the same way, employees will find it hard to follow certain safety rules and procedures if they are consistently (certain) rewarded by an immediate (soon) timesaving that achieves extra production (positive) by behaving unsafely. What would you do, for example, if you were faced with a ten to fifteen minute period to put on the correct clothing and equipment to enter a mandatory PPE area to read a gauge that takes only 10 seconds?

In some instances, the actual workflow process also reinforces peoples' unsafe behavior. By way of example, supermarket 'shelf' stackers were required to replenish freezers with frozen foods. Warehousemen fully load five-foot high, three-sided meshed trolleys with large cartons of frozen foods. The stackers push the trolleys from the warehouse and transfer the individual packages into the appropriate freezer. Unfortunately, no provision was made for the storage or disposal of empty cartons. To ensure customer access to the surrounding freezers, while also avoiding the creation of tripping hazards, the stackers were forced to place all the empty cartons on the top of the trolley at the rear. This particular unsafe behavior had the potential to injure both customers and stackers due to the trolley tipping over when the last carton was removed from the trolley because of the additional, unbalanced, weight. Nevertheless, it soon became part of the stackers normal way of working, because their behavior was always (certain) reinforced immediately (soon) by getting their job done (positive) to maintain sales volume.

Unsafe behavior is sometimes further reinforced by line managers turning a blind-eye, or actively encouraging employees to take short-cuts for the sake of production. Unfortunately, this has negative effects that are not always immediately apparent: First, employees learn that unsafe behavior pays; Second, it wastes resources as the very behaviors that companies spend a lot of time, money and effort trying to eradicate are reinforced; and third, by condoning unsafe behavior, line managers are transmitting conflicting messages that undermines employee's confidence in the whole of management's commitment to safety.

How Do We Stop People Behaving Unsafely? Why not engineer out hazards?

Eliminating hazards by engineering them out or introducing physical controls can be an effective way of limiting the potential for unsafe behavior. While successful in many instances, it does not always work, simply because people have the capacity to behave unsafely and override any engineering controls.

For example, in attempts to reduce the number of fatalities associated with quarry transport, companies install belt conveyers to replace vehicles as the main haulage system for transporting extracted minerals. To overcome major operational problems associated with these conveyance systems (e.g. the spillage of minerals at transfer points from the belt), engineer’s design and install belt scrapers to minimize mineral build-ups at the pulleys to reduce belt distortion. Despite these precautions, materials often build-up at the nip point between belt and pulley. When this occurs, it is not unusual to find operators removing the guards while the belt is still in operation to clear the material build-up. Others are known to attempt to clear the moving pulleys with iron bars or shovels. In both cases there is a high risk of the tools becoming caught in the nip points of a pulley, and drawing the operator in with serious consequences.

Clearly, despite the presence of the machine guards, operators often believe that the consequences of behaving unsafely will be more than repaid by continued production. This illustrates the point that many engineering solutions tend to be reliant on peoples' 'rule following' behavior (e.g. stopping the machinery before removing guards) but people still have the capacity to ignore them and behave unsafely.

Thus, although engineering solutions have a strong place in safety management, they cannot be relied upon.

How Do We Stop People Behaving Unsafely? Why not change people’s attitudes?

Comments on accident reports often say 'So and So should take more care. With better attitudes and safety awareness, this accident would not have happened'. Where this occurs, attempts to change unsafe behavior usually hinge upon the belief that attitudes determine behavior (Indeed, this is a very common opinion amongst many safety professionals). Remedies tend to rely on publicity campaigns and safety training to bring about changes in people’s attitudes, which in turn is expected to change people’s behavior. Although positive safety attitudes are important and very desirable, the link from attitude change to behavior change is very weak. This can be explained by the fact that a single attitude comprises of at least three components: thinking (cognitive), feeling (emotional), and the intention to act on it (commitment). Additionally, a single attitude is usually linked with a set of other related attitudes. Logic dictates, therefore, that attempts at attitude change must target each individual component of each individual attitude, for each single employee. In practical terms this is nigh on impossible.

Fortunately, the link from behavior change to attitude change is much stronger. If people consciously change their behavior, they also tend to re-adjust their associated attitudes and belief systems to fit the new behavior. This occurs because people try to reduce any tension caused by a mismatch between their behavior and attitudes. Behavior change, therefore, tends to lead to new belief and attitude systems that buttress the new set of behaviors. This was demonstrated in a manufacturing plant, where the safety climate of the company was measured before introducing behavioral safety. Eighteen months later, the company's safety climate was re-measured. Positive changes occurred in six of seven topic areas (Cooper & Phillips, 2004).

An additional factor that enhances attitude change by focusing on behavior is the positive reinforcement brought about by peer pressure. Psychologists have known for some time that group membership demands conformity to the groups' behavioral and attitudinal 'norms'. If a workgroup adopts the 'norm' that 'thinking and behaving safely' is best for all concerned, the group as a whole will tend to apply social 'sanctions' to the individual who deviates from this norm and behaves unsafely. If people wish to remain a part of the social fabric of the workgroup, they soon revert back to the safety norm and behave safely. Importantly, this illustrates the point that workgroups will adopt a collective definition of those behaviors, work practices or tasks that are considered to be risky (Cooper, 1997). This fact lies at the very heart of behavioral safety, simply because its essence is to help workgroups positively redefine their own safety related 'norms.

How Do We Stop People Behaving Unsafely? Why not punish people until they behave safely?

Some approaches to safety management are heavily reliant on the use of authority, fear and punishment (i.e. if you do not behave in a safe manner at work you could be reprimanded, fined or even dismissed). These approaches emphasize the use of discipline and punishment to discourage unsafe behavior, while safe behavior is largely ignored. This often results in the opposite of that intended (e.g. accident or near-miss incidents are not reported for fear of sanctions). Although the judicious use of discipline and punishment can have the intended effects, more often than not it doesn't. The reason for this is quite simple: The effectiveness of punishment is dependent upon its consistency. It only works if is given immediately, and every single time an unsafe behavior occurs. It is self-evident that punishing someone every time they behave unsafely is a very difficult thing to do, simply because they will not always be seen to do so by those in authority. This means those soon, certain and positive reinforcers gained from behaving unsafely will tend to outweigh any uncertain, late, negative reinforcers received from inconsistent punishment.

Thus, although punishing those who deliberately put other people at risk is a valid option, punishing people for everyday infractions of safety rules (e.g. not wearing a hard hat) is a very difficult thing to do consistently and does not address the underlying problems (e.g. the hard hat is uncomfortable or gets in the way of doing the job safely). In essence, therefore, punishment should be reserved: [1] to those who deliberately put other people at risk; and [2] only after the organization has done everything in its power to create the safest working environment, provide the most comfortable protective equipment and a persons' unsafe behavior is a consistent, willful act aimed at flouting authority.

How can we stop unsafe behavior? Why not praise people for behaving safely?

So how can line management ensure that the reinforcers for working safely outweigh those for working unsafely? It is a fact that most people tend to respond more to praise and social approval than any other factor. Think if you will, of people smoking their first cigarette. This normally occurs during the teenage years because it is seen as the 'thing' to do. Although the cigarette smoke may taste foul and cause severe coughing, people will continue to suffer the discomfort, if the cigarette smoking behavior meets with their peer group's approval. Likewise, some people may not use PPE or follow a procedure at work because of their colleague’s disapproval: e.g. it goes against the workgroup's macho image.

It makes sense, therefore, to make use of this phenomenon and praise people for behaving safely (something very rarely done) to bring about the required changes (Incentive and reward schemes reflect this principle). Crucially, the effect of this is to explicitly link the desired safe behavior to the praise received. Once the required behavior pattern starts to become established, the timing and frequency of the praise and social approval can be reduced over a period of time: i.e. it doesn't need to be given immediately and every single time that someone is seen to be behaving safely. Additional benefits include the strengthening of a positive safety culture due to increased trust and confidence between line managers and the workforce.

Thus, positive praise coupled with constructive feedback, tends to eliminate unsafe behavior.

What next?

We know that focusing on people's safety behavior will bring about the desired changes and that attitude changes follow behavioral changes. We know that social approval and encouragement can bring about positive changes in safety 'norms'. We also know the workforce is best placed to redefine their safety 'norms, as they control their own behavior. It follows, therefore, that any safety improvement initiative which relies almost exclusively on line management's efforts, is less likely to be as successful as one that empowers and enables the workforce itself.

Accordingly, behavioral safety approaches are very much driven and shaped by the workforce, in conjunction with line management. In this way, the workforce is given responsibility and authority for identifying, defining and monitoring their own safe and unsafe behaviors, as well as setting their own 'safety improvement' targets. As a result, workgroups are able to redefine their own safety related 'norms' in an enabling atmosphere. Line management facilitate this process by providing the necessary resources and support to encourage 'employee ownership of safety', while also stressing that no individual will be identified or disciplined as a result of the monitoring. In this way a 'blame free' pro-active safety culture is created that is so vital for long-term success.

Have we provided insights into many of the issues you face? If so, perhaps now is the time for your company to consider behavioral safety as the way forward. But you are wondering if it will work.....

Does it work?

Because the behavioral approach differs considerably from traditional ways of improving safety, a question commonly asked is 'Do these ideas work in practice?' Overwhelmingly, the answer is yes! Researchers from around the globe have consistently reported positive changes in both safety behavior and accident rates, regardless of the industrial sector or company size (See references section). These include studies conducted in construction, mining, engineering, bakeries, food processing, manufacturing, oil & gas, shipbuilding and others (See the references section particularly the reviews by McAfee &Winn, 1989 & Sulzer-Azeroff et al., 1994, Grindle et al, 2000). Positive results have also been obtained by the author and colleagues over the last decade in many sectors of the UK, Irish and US economies, such as Construction (Duff et al., 1993), Manufacturing (Cooper et al., 1994), Chemicals, Pharmaceuticals, Paper, Foods, Steel, Paints and Offshore Oil & Gas. Typical results include:

1. 40-75 percent reductions in accident rates and accident costs year on year
2. 20-30 percent improvements in safety behavior year on year
3. Greater workforce involvement in safety
4. Better communications between management and the workforce
5. Ongoing improvements to Safety Management Systems
6. Improved Safety Climates
7. Greater 'ownership' of safety by the workforce
8. More positive attitudes towards safety
9. Greater individual acceptance of responsibility for safety.

Thus Behavioral Safety has a lot to offer to the world of work, although it must be stressed that it is not a panacea to cure all ills.


Cooper, M.D.(1998) Improving Safety Culture: A Practical Guide. J. Wiley & Sons, Chichester.

Cooper, M.D. & Phillips, R.A.(2004) Exploratory Analysis of the Safety Climate and Safety Behavior Relationship. Journal of Safety Research, 35, 497-512.

Cooper, M.D., Phillips, R.A.,Sutherland, V.J. & Makin, P.J. (1994) 'Reducing accidents using goal- setting and feedback: A field study'. Journal of Occupational & Organisational Psychology, Vol 67, 219-40.

Duff, A.R., Robertson, I.T., Cooper, M.D. & Phillips, R.A. (1993) 'Improving safety on construction sites by changing personnel behavior'. H.M.S.O. Report Series CRR51/93:H.M.S.O. ISBN 011 882 1482; London.

McAfee, R.B. & Winn, A.R. (1989) 'The use of incentives/feedback to enhance workplace safety: A critique of the literature'. Journal of Safety Research, Vol 20, 7-19.

Sulzer-Azeroff, B., Harris, T.C.,& Blake- McCann, K. (1994) 'Beyond Training: Organizational Performance Management Techniques'. Occupational Medicine: State of the Art Reviews, Vol 9(2), 321-339.