Notice: Passwords are now case-sensitive

Remember Me
Register a new account
Forgot your password?



Community Requests

Moore: Accident Curve and Natural Disasters

By James Moore

Monday, September 18, 2017 | 822 | 0 | min read

The accident curve measures the accident rate lessening per unit of work time. The concept has been debated over time. A search far and wide located a 1916 text that actually addressed this issue very well.

James L. Moore

James Moore

The learning curve measures how long it takes for a person to learn a new task. The accident curve measures how many accidents happen in the first few hours someone works with a new machine or performs a new task.  

I performed an informal study on the accident rates of claims that I was handling post-Hurricane Hugo (September 1989). After two months, the claim numbers spiked and then returned to normal.

Most of the injured workers were new to the construction industry. They were involved with the hurricane recovery efforts for many months.    

We have all heard the expression “90% of the accidents occur the first time someone uses a tool/machine” — or the first time a person performs a completely new task.  

Henderson's law is Cn = C1n-a.


  • C1 is the cost of the first unit of production.
  • Cn is the cost of the n-th unit of production.
  • n is the cumulative volume of production.
  • a is the elasticity of cost with regard to output.

A few classes I attended to complete my actuarial degree covered the learning curve heavily. The application centered around production. Safety did not seem to be an issue.

However, the learning curve is a the crux of a safe workplace. Hands-on training reduces workers' comp accidents or their severity by exposing the workers to a learning curve with a person or persons who have long since handled the type of new task.

The actual formula uses logarithms to estimated the time it will take someone to complete the task. A very astute scientist derived an accident curve from the formula.  

However, he derived the accident curve from actual hands-on measurement in manufacturing plants. You will need to refer to the bottom page 127 of this old but still relevant text.

The text is included below justifying the accident curve consideration. The studies were conducted on very heavy labor iron and steel plants. 

Accidents and inexperience the high risk of beginners

It is generally recognized that inexperience is one of the most important elements in the causation of industrial accidents. What has not been sufficiently realized is the close parallelism of the rates of accidents and of inexperience.

The industrial risks of beginners have been shown to be high. Interesting testimony on this subject is given in the iron and steel report already quoted.

A table giving the incidence of accidents to press hands beginning work on the machine shows that on the first day of employment:

  • More than 5 times as many were injured as during all the rest of the first week.
  • 26 times as many as from the second week to the end of the first month.
  • 100 times as many as from the second through the sixth month.

One can presume the figures indicate a six-month employee is 100 times less likely to have an accident as the first-day employee.

Another table covering longer periods of employment shows the accident frequency rates per 1,000 workers at a steel plant falling from the exceedingly high rate of 111.3 per 1,000 in the first six months through increasing lengths of service, to 19.7 for from five to 10 years and to zero after 15 years. 

The book did not mention the level of training in these cases. One would have to guess that it was strictly on-the-job-training.

This blog post is provided by James Moore, AIC, MBA, ChFC, ARM, and is republished with permission from J&L Risk Management Consultants. Visit the full website at www.cutcompcosts.com.


Be the first to comment.

Related Articles