In the 1930’s, William Herbert Heinrich, an employee working for Traveler’s Insurance Company, published  groundbreaking theories about safety and health in the workplace. One such theory became known as Heinrich’s Law:
that in a workplace, for every accident that causes a major injury, there are 29 accidents that cause minor injuries and 300 accidents that cause no injuries.
Because many accidents share common root causes, addressing more commonplace accidents that cause no injuries can prevent accidents that cause injuries.
Heinrich’s work is claimed as the basis for the theory of behavior-based safety by some experts of this field, which holds that as many as 95% of all workplace accidents are caused by unsafe acts. Heinrich came to this conclusion after reviewing thousands of accident reports completed by supervisors, who generally blamed workers for causing accidents without conducting detailed investigations into the root causes.
While Heinrich’s figure that 88% of all workplace accidents and injuries/illnesses are caused by man-failure is perhaps his most oft-cited conclusion, his book actually encouraged employers to control hazards, not merely focus on worker behaviors.
No matter how strongly the statistical records emphasize personal faults or how imperatively the need for educational activity is shown, no safety procedure is complete or satisfactory that does not provide for the . . . correction or elimination of . . . physical hazards,
Heinrich wrote in his book.
Heinrich’s theories were considered sacrosanct, until Fred Manuele, President of Hazards Limited, a consulting firm, recently published  an article in the ASSE periodical – Professional Safety challenging the validity of Heinrich’s Law. The challenge specifically targets this premise:
focusing on incident frequency reduction will equivalently achieve severity reduction.
What Mr. Manuele found is that if you manage the small incidents effectively, the small incident rate improves, but the major accident rate stays the same, or even slightly increases.