Safety Myths and Misconceptions

Safety Myths & Misconceptions

Republished in memory of the late George Robotham (see all of his articles here) and his mentor Geoff McDonald who also recently passed away (

'The Myth' photo (c) 2009, Ata Ur Rehman - license:

Note:- Brisbane based OHS consultant, Geoff McDonald has considerably influenced my thinking in the preparation of the following. Because they go to work in Australia, 10 people per hour, 24 hours a day, 7 days a week, 52 weeks a year, have their lives permanently altered (Geoff McDonald).


One important factor that influences how OHS is managed is the attitudes and pre-conceptions of those leading the safety charge. This paper explores beliefs, philosophies, concepts and attitudes and suggests some common ideas may be incorrect or unhelpful, that is they may be myths and misconceptions.

Essential concepts

Damage to people at work has a number of adverse outcomes:

1.       Financial loss to employer, worker and community

2.       Pain and suffering

3.       Dislocation of lives

4.       Permanence of death

5.       Damage to people from work falls naturally into one of three Classes.

6.       Class I damage permanently alters the person’s life and subdivides into

·         fatal

·         non-fatal

1.       Class II damage temporarily alters the person’s life( Many of the lost time injuries)

2.       Class III damage temporarily inconveniences the person’s life, minor injuries (Geoff McDonald & Associates)

Whilst recognising the unique nature and devastating effect of Class 1 fatal personal damage the impact of the massive amount of Class 1 non-fatal personal damage is largely unrecognised. We get to hear about some of the fatalities and rarely get to hear about the non-fatal Class 1 personal damage.

All personal damage occurrences will have People essential factors, Machine essential factors and Environment essential factors. Essential factors, commonly referred to as causes, are those without which the final damage could not have occurred. Cause is an emotionally laden term, infers blame and should not be used in investigations. The term accident is also emotive, infers blame for many people and is better replaced with the term personal damage occurrence.

Suggested myths and misconceptions

People cause accidents

We would not suggest that people are not essential in personal damage occurrences (Accidents) but the people cause accidents myth and misconception is often used as an excuse for not carrying out positive action. What often happens is we blame the person and forget about making positive changes to the machine and the environment. There are few occasions when it is appropriate to blame the person for their past actions, this is only appropriate when the blame leads to change in the future.

The people cause accidents philosophy has been reinforced in a number of ways over the years.

Heinrich-Although this belief has been part of our culture for centuries, it received official sanction in the writings of Heinrich, widely held to be the father of the industrial safety movement in the 1930’s.

His domino theory whereby unsafe acts, unsafe conditions, errors and hazards combine to produce incidents has tended to focus on the person to blame and has been a serious impediment to meaningful progress.

Legal system-This reflects the belief that people cause accidents. The system is seen by many to be nothing more than a crime and punishment system, where people are held to blame and punished accordingly. No other factors than people’s actions are given consideration when judgements are made in damages claims arising from motor vehicle accidents.

Insurance industry-Closely tied in to the legal system, seeks to identify some person to blame and pursue through legal channels for any claim.

News media-Media scream driver error in motor vehicle incidents, they scream pilot error in aviation incidents without taking account of the other multitude of essential factors.

Published studies-Many published studies will have you believe 90% of accidents are caused by human error. The reality is all personal damage occurrences will have people essential factors and machine and environment essential factors.

The main aim of safety activities is to prevent accidents

Certainly safety activities aim to prevent personal damage occurrences. However we must take one step further by also seeking to minimise and control damage. A classic example being the wearing of seat belts and fitting R.O.P.S. to tractors.

Look after the pence and the pounds will look after themselves

There is a belief in safety that if you bring controls to bear on all minor injuries then the Lost Time Injuries will look after themselves. This belief has mis-directed effort with the result that inordinate effort is directed at minor incidents that have little potential for more serious damage. Certainly we should prevent minor incidents but remember to concentrate our efforts where we get the best results. The Pareto Effect says 20% of incidents will give 80% of damage. This 20% must be identified and concentrated upon. What you must to is direct your safety efforts where you will get the biggest bang for your buck.

In Managing Major Hazards Professor Andrew Hopkins outlines how a focus on Lost Time Injuries led to insufficient emphasis on high risk events. Papers are emerging questioning the wisdom of Zero Harm approaches to safety.

It cannot happen to me

There is a need for each and everyone of us to subscribe to this theory, for the sake of our own psychological well-being and to be able to cope with situations outside our control. This belief is often no more than an excuse for taking no action. Often you will wonder why the silly bugger did what they did, sometimes it is because of this belief.

Punishing wrongdoers

I am not saying we should not punish people who do the wrong thing in safety. I am saying that the fact that we do punish wrongdoers will often lead to highly imaginative efforts to avoid punishment and thus make things harder. The history of the safety movement records numerous cases of punishing the wrongdoers not being effective. We should seriously consider the full range of options rather than making hasty decisions to punish the wrongdoers.

W.A.S.P. ethic  (White Anglo Saxon Protestant)

This work ethic had its origins in the great religious upheaval know at the Reformation. The ethics emphasis is just reward for effort, conversely people who are hurt in accidents are receiving their just reward for lack of effort. The W.A.S.P. may side-track our prevention efforts.

Displacement activities

A displacement activity is something we do, something we put a lot of effort into but which there is no valid reason for doing it. Examination of the history of the industrial safety movement will reveal many examples of displacement activities. Zero Harm and some aspects of Behaviour Based Safety strike me as examples of displacement activities. The use of many safety posters and blood and guts accident photographs would also be included.

Lost Time Injury Frequency Rate is a valid and reliable measure of safety performance

I have personal experience with a company that aggressively drove down L.T.I.F.R. to a fraction of its original rate in a space of about 2 years yet killed 11 people in one incident.

The Lost Time Injury Frequency Rate predominates discussions about safety performance. How can a company be proud of a decrease of L.T.I.F.R. from 60 to 10 if there have been 2 fatalities and 1 case of paraplegia amongst the lost time injuries? The L.T.I.F.R. trivializes serious personal damage and is a totally inappropriate measure of safety performance.(Refer to the paper on this topic under articles on

Managers understand learning needs

Every task that needs to be done by people must be done

With appropriate consideration for people, for the community and for the Environment (Competency-Based Learning)

Detailed task analysis must take place to recognize the safety competencies required to perform all tasks (including supervisory) where gaps exist between required competencies and current competencies appropriate learning may be the most appropriate solution. After people attend learning exercises the supervisor should develop a plan, in association with the trainee to implement the lessons learnt. After learning projects should receive consideration. A specific program of learning needs analysis is required to identify learning needs, do not rely on gut feel.

Risk Assessment

Notwithstanding the popularity of risk assessment techniques there are some limitations to the techniques that need to be realised. I have always been of the view that what you do to control risk as a result of a risk assessment exercise is more important than the risk rating. Placing too much emphasis on comparison of risk ratings will lead to inappropriate priorities. Risk assessment exercises are often subjective. When it comes to developing controls  I find Haddon’s  10 countermeasures more effective than the hierarchy of controls.

Safety Procedures are the answer

The commonest mistake I have seen with safety management systems is the development of extensive safety procedures that the workers do not know about, care about or use. The procedures sit on the supervisor’s bookcase or a computer program and are rarely referred to. The job safety analysis technique must be used to develop safe working procedures and involvement of the workforce is crucial. If your safe working procedures are over 2 pages in length worry about whether they will ever be used. Use flow-charts, pictures and diagrams in your safe working procedures and base them on a very basic level of English. The K.I.S.S. principle applies.

Critical incidents or near-misses are well reported

Critical incidents (near misses, sometimes referred to as near hits) occur regularly in organisations but are not routinely reported for a number of quite valid reasons.   Critical incidents must be surfaced through an organised process. Critical incident interviewers and observers must be trained and they should spend some time in the organisation identifying critical incidents. Exploring why critical incidents occur will provide significant insight to guide the safety management system (Refer to the paper “Practical Application of the Critical Incident Recall Process” by this author)

Analysing enterprise accident data is a good idea

Unless you are a very big organization only limited insight into future class 1 personal damage will be gleaned from analysis of enterprise experience. Taxonomies of industry experience can be a powerful tool. The development of standardized, industry reporting systems is underutilized.


It is suggested some common approaches to OHS may be myths and misconceptions. The situation is probably best summed up by an ex-manager of mine who says the biggest problem with safety is that managers and safety professionals often engage in acts of public masturbation.


A few responses from the LinkedIn Group Work Health Safety leadership

A good account of myths and misconceptions. I am also a supporter of the research work of Geoff McDonald and have posted some information with reference to Geoff’s work of over 40 years.  This “leadership group” could benefit from reading Geoff’s book “SAFE FROM” – Information Organisers which is unpublished but I am sure Geoff would be pleased to send copies to those interested in improving how we can manage Safety in all aspects of society; be it at work; home; school; travel; recreation and sport.

In safety, the word ‘cause’ is strongly associated with ‘blame’. The Australian Concise Oxford Dictionary gives the following:
blame. v. tr. 1. assign fault or responsibility to. 2. (foll. by on) assign the responsibility for (an error or wrong) to a person etc. n. 1. responsibility for a bad result; culpability (shared the blame equally; put the blame on the bad weather) 2. the act of blaming or attributing responsibility; censure (she got all the blame)
Experimental research described the same car crash with progressively more emotionally upsetting outcomes from car damage only to injury of an adult to death of a young child. As the emotional severity of the outcome increased, so did people’s perception of the culpability or blameworthiness of the driver.

‘Cause’ has long been a difficult question for scientists and writers alike.

Heinrich made a valuable contribution by bringing ‘accidents’ out from being only vaguely understood and provided his ‘domino’ model which gave a causation model which led him to conclude that the ‘Cause of Accidents’ was:

Human Failings 88% of cases
Mechanical Failure 10% of cases
Acts of God 2% of cases

This work of Heinrich came before 1950. In 1973, a group from Indiana University gave the ‘causes’ of road ‘accidents’:
Human 75% of cases
Environment 15% of cases
Vehicular 6% of cases

McDonald (G.) analysed 520 permanently life altering tractors cases to identify factors essential to the final damage (over 300 were fatal) and found:
Human Behaviour in 98% – up to 9 per case
Tractor Design in 95% – up to 6 per case
Surroundings (Environment) in 80% – up to 5 per case

The number of essential factors was under-recognised because of loss of information through death, and, un some cases, insufficiently detailed reports. Further, McDonald analysed the last 420 cases by himself and identified more behaviour factors than design factors than in the first 100 where two other experienced tractor drivers helped with the analysis. The method of identifying essential factors has been further developed and would have identified more factors, had it been available and used.

The best conclusion is that behaviour factors, design factors and environmental factors were each essential factors in 100% of cases. Since all essential factors are equally important in causation, there is no thinking basis for elevating one above the others by nominating it as a ‘cause’. Essential factors differ in their controllability and that is where the major decision making should concentrate. What is the most effective factor to control in the short term, in the long term, in the limited sphere of influence and in the wider sphere of influence?

Identification of a large number of essential factors gives many more options for control. Experience has shown that when the range of factors is identified, people willingly volunteer responsibility for factors ‘down to them’ and start discussing how they can control those factors including their own behaviour. They do not wish to be unimportant or irrelevant and usually wish to be carrying their own share of the efforts. This experience reflects years of use of the essential factors methods.

The use or the term ‘cause’ fits well with the Egocentric Model (III From) and the Feeling Function (IV From) while Essential Factors fit well with the Ergonomic Model (III To) and the Thinking Function (IV To).
The use of the term ‘cause’ because of its strong link to ‘blame’ and thereby to ‘fault’, ‘error’, ‘wrong’, ‘bad’, ‘censure’ and the like, takes the feeling judgement function beyond its normal range by helping produce high levels of emotion which will more strongly distort objectivity when there is a strong need for clear thinking.

The only basis for elevating an essential factor to a higher status and labelling it ‘cause’ is by using the feeling judgement function which seeks to make a good judgement – with the ‘goodness’ being according to the values of the person making the judgement. No particular skills or knowledge are required to make such judgements, only strong feelings. The ‘basic, root, fundamental, prime and ultimate cause’ usually suggests a cause of greater importance or requiring greater skill to determine. It is usually the essential factor with the strongest emotional appeal to the most dominant influence involved in the analysis and assessment. ‘Proximal’ and ‘distal’ represent proximity to or distance from the damaging energy exchange.

Some argue that multiple causes is similar to essential factors but experience shows that fewer causes are identified as the feelings remain blinkering.

Certainly there are situations where a strong emotional response is almost inevitable. Allow your humanness to feel the emotion and then come to rest. Then use your professional skills, collect data carefully and thoroughly, seek understanding of what at first appears unreasonable and what mystifies you and only after that build and test your hypotheses as the essential factors reveal themselves.

Geoff McDonald is the clearest thinker about safety I have come across.His published works make a significant contribution.Geoff is a great critic of traditional approaches to safety, for very good reasons.He has had a profound effect on how I view and approach safety

I agree entirely with your praises about our friend and mentor Geoff McDonald.
Like you, I have also known Geoff for over 40 years and have used his teachings and practical approaches in my Safety Engineering and Managerial work in many companies and industries.
I attended one of the first of Geoff’s Analysis Reference Tree Trunk method of accident investigation; and I have used his Essential Factors methodology many times. Like you, I also have many examples where successful prevention solutions have been developed and implemented.

Geoff, I think it would be a great idea if you could share your “epitaph” SAFE FROM with the members of our LinkedIN group which presently numbers about 7,000 members.
And yes George; even a little difference can make a big contribution.
“What we do to improve (safety in workplaces, my emphasis) is like a drop in the ocean; but if that drop was missing the ocean would not be the same”! (Mother Teresa)

Essential Factors Method of Investigating Damage Occurrences (Ref: G. McDonald)
From the above discussions “Cause to Essential Factors”, the following personal example may serve to illustrate the difference between what we see as “Cause” and what can be effectively done to control damaging occurrence.
Consider this Damaging Occurrence (Accident/Incident?) scenario:

A Council worker was engaged in driving a tractor pulling a rotary slasher type of mower which was hung on the hydraulic hitch and driven from the power take off (PTO) at the rear of the tractor. Just before lunch he decided to grease the mower bearings at each end of the two horizontal levelling rollers; and parked the unit on level gravel ground near the lunch room. The 4 grease nipples on the mower were oriented exactly horizontally, except one, which had perhaps been replaced at some stage by one nipple angled downwards at about 45 degrees. This nipple was a problem because the grease gun had a rigid head, so that the worker found it necessary to raise the mower attachment to gain sufficient clearance to use the grease gun. This in turn required him to start the tractor engine, which he did while standing next to the tractor on the ground. With the mower raised he applied grease through the nipple, lowered the mower and reached between the gear levers of the tractor to pull the stop lever of the diesel engine on the opposite lower side. Unfortunately, the arm movement knocked one of the gear levers (H/L ratio) into gear and, as the other lever was parked in first gear, the tractor moved forwards. Now although the worker had applied the park brakes they were only working on one side because of a bracket failure which prevented the brake pedal on the other side of the tractor from where he was standing being held on. The result was that the tractor moved in a circle, knocking down and then running over the worker with its large rear wheel followed by the mower. The tractor continued in a circle and ran over him twice before one of the other workers realised what was happening and stopped the tractor only a short distance before it would have run over the victim a third time. The worker died in hospital two weeks after the damage occurrence and was able to recount exactly what happened before he died.
I investigated this fatal occurrence using the Essential Factors methodology and numerous factors were identified, including at least these:

1.The angled grease nipple,
2.The rigid head on the grease gun,
3.The failed lock bracket on the park brake,
4.Stopping the diesel engine by pulling out the stop lever,
5.The need to reach through the gear levers to stop the engine,
6.The operator was able to do this while standing on ground in front of the large rear wheel,
7.The possibility of knocking the tractor into gear (H/L ratio) with the engine running and the drive lever in first gear,
8.The worker being alone so that he was tempted to start and stop the engine from the ground and so that the tractor could run over him twice.
9.In their earlier years, tractors were not designed and built with seat interlocks such that it is only possible to start them if the operator is sitting in the seat; and so that they will not keep running if that person leaves the seat or perhaps even falls from the seat.
10.Also, early tractors were not fitted with interlocks on each of the gear levers so that the engine could not be started unless the levers were in neutral position.

The next step in the investigation was to identify which of the above factors were “Essential” and which were contributory at the time.
Perhaps the readers of this “Real” case study would like to attempt to identify the Essential factors as by doing this we can prove that “Risk can be eliminated” or at least, reduced.

Good example, I am also a fan of the Geoff McDonald Analysis Reference Tree Trunk method of investigation.I have looked at most of the other methods but keep coming back to A.R.T.T.
Cause is an emotionally laden term that infers blame,I find using the essential factors methodology, guided by the A.R.T.T. model and worksheet alllows me to better understand the event.This understanding is essential to develop relevant controls.
Geoff says Brisbane based OHS consultants, Intersafe run very good training on A.R.T.T.
For you cynical bastards out there please note that I have no commercial ties with Geoff McDonald or Intersafe
I promote A.R.T.T. simply because it represents best practice in what I believe is a space that could do with some improvement

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.