Javascript DHTML Drop Down Menu Powered by



(re safety differently)



Following his analysis of 75,000 industrial accident reports (see more here), Heinrich (1941) put forward the following amongst his many critical findings...


  • 2% of accidents are “unpreventable” (p.19).
  • 10%  should be “charged to defective or dangerous physical or mechanical conditions”; and
  • 88% “are caused primarily by the unsafe acts of persons” (p.20).


On page 19., he says “man failure causes or permits” unsafe acts and conditions (i.e. the 88% & 10% above). To be precise, he found that 73,500 of the accidents in his study had “the same causes” (p.390); i.e. “man failure”.


Believing that “the complete ideal programme of accident prevention awaits the millennium” (p.8) and that industry was “not yet receptive enough to proceed with the next logical step – accident psychology” (p.283), Heinrich was sometimes blunt. However, he understood terms other than ‘man failure’ and ‘unsafe acts’ and soon substitutes them for error, violation and mistake (p.390).  


Later, discussing his 300:29:1 ratio, he says “the realization that causes were alike for the entire unit made it no longer necessary to await the one serious accident before effecting the necessary correction for the entire group so that the eventual serious accident could be avoided” (p.384).   


Academics and writers now use the term ‘common cause hypothesis’ in respect of the finding that industrial accidents have alike causes (academics actually refer to common causal pathways). Many have attempted to reject the finding because, not least, it can present an obstacle to certain approaches to causation. In the main, ‘challenges’ are levied when researchers find ratios that are different to Heinrich’s estimated 300:29:1. Of course, the belief that a fixed ratio should exist is a fallacy.


Interim Comment:

The common cause hypothesis has never been refuted and modern research continues to find support for it (e.g. Davies et al 2003). Nonetheless, some (e.g. Dekker et al 2008 re resilience engineering) have claimed that it might be valid but, only in respect of near-miss and minor injury accidents. However, analysis of that claim reveals a degree of confusion and inconsistency. Dekker et al (2008) rely on Hollnagel yet, Dekker (2014) ( i.e. safety differently) offers that the common cause hypothesis might only be wrong “at 10-7 and beyond”. (readers will find more on the 10-7 aspect and resilience as they progress through these discussions).



The Institute of Industrial Accident Investigators. All rights reserved.