Javascript DHTML Drop Down Menu Powered by



(re safety differently ADQ3. a. – b.)



3. On page 21. (under ‘Change orders’), we see how three fatalities on the company’s sites in November 2013 and July and October 2014 are behind its journey.


John Green (Laing O’Rourke’ H&S Director) says the sites showed “no early warning signs of a lax safety culture” nor did they “send up the early warning flags that ‘Heinrich’s triangle’ would suggest”. The article also says “the new gospel has its own vocabulary” and Green then refers to the “new safety paradigm”. 




a). Why did Heinrich produce his triangle?

b). Does the triangle suggest such “early warning flags”?

c). According to John Green, what is the “new safety paradigm”?



This is actually the second of two references by John Green to Heinrich’s ‘triangle’.

Note: the triangle in question first appeared in a 1929 presentation by Heinrich.


In his seminal work, “Industrial Accident Prevention – A Scientific Approach”, Heinrich explains the findings of his analysis of 75,000 accidents.


On p.26. he says “analysis proves that for every mishap resulting in an injury there are many other similar accidents that cause no injuries whatever” (1941 2nd Ed).


He continues, “from data now available concerning the frequency of potential-injury accidents, it is estimated that in a unit group of 330 accidents of the same kind, 300 result in no injuries, 29 in minor injuries, and 1 in a major or lost time injury”.


To graphically portray things, he depicted a ‘triangle’ with the number 300 at its base, the number 29 at its mid point and the number 1 at its top (1941. p.27.).



i).  The triangle represents a unit group of 330 accidents of the same kind.

ii). Heinrich draws attention to the fact that his figures are estimated.



Obviously, “the laws of chance” (p.26.) are such that “major injury may result from the very first accident or from any other accident in the group” (p.27.). Of course, “there are certain types of accident where the probability of serious injury may vary in accordance with circumstances” (p.28.). Later, he reminds us again of the massive role played by “luck” (p.342.).


Note: He emphasises that chance and luck are always involved.  



Having highlighted that his figures were estimated, he emphasises further by saying the actual number of near-miss accidents is a “nebulous quantity” (p.30.) and the 10:1 ratio is “ultraconservative” (p.34.). Indeed, he describes accidents within his study where, for instance, the ratios were 1800:1, 1500:1 and ‘several 1000 to 1” (p.33.).



i). He has now used the term nebulous quantity and the word ultraconservative.

ii). Triangles produced from his actual data would each have had different ratios.



Describing a particular accident involving a machine, Heinrich says the “defect had existed throughout [its] life. Over...several years, each revolution...had exposed the operator to injury. Finally, the exact balance of variable circumstance occurred, and an injury resulted” (p.31.).


Note: He now introduces the term variable circumstance.



Interim Response:

The main aim of the triangle was to graphically portray the fact that far more near-miss accidents occur than do loss, damage, or injury producing ones.


Words like ultraconservative, luck, probability, nebulous quantity, variable circumstances, estimated and chance are hardly the words of certainty.  


Clearly, Heinrich neither claimed nor suggested that a triangle could predict (let alone send up “early warning flags” in respect of) some random future accident. Evidence in support of that has been with us for 88 years.  

Similarly, reference to the absence of an early ‘safety culture’ warning is also puzzling since safety climate (i.e. the measure of safety culture) relates only to the perceptions of certain “individuals or work groups at a point in time” (HSE 2009).


General Comment:

In the same way that Heinrich never claimed (nor could) that there was any predictive capability in a triangle, he neither argued nor suggested that 300:1 (or any other ratio) should be taken as a universal norm (recall his 1800:1, 1500:1 and several 1000 to 1. ratios).


Neither F.E. Bird (1974), Bird & Loftus (1976) nor the UK Health & Safety Executive (1997), purely for instance, made any form of criticism despite their own studies having generated six different ratios between them.


Nonetheless, a number of academics have ‘challenged’ Heinrich’s 'common cause hypothesis' based on an apparently ‘different’ ratio emerging from UK Health & Safety Executive statistics. Since two of the academics (i.e. Dekker & Hollnagel) are referenced in the safety differently article, the ‘challenge’ should be aired and analysed as part of the overall discussions here in the interests of balance.

Click here for comment on the first part of the challenge.









The Institute of Industrial Accident Investigators. All rights reserved.