By louis
Published: October 27, 2007
There was an intersection that had a thirty-percent higher than average accident rate. The city held a meeting and decided this was unacceptable and set a goal to reduce accidents by 10% a year for three-years. They announced the goal with a great deal of fanfare and sent out a memo to all departments.

The next year the accident rate was the same as the year before. Perplexed, they decided perhaps the goal was not met because there was no incentive attached. They sent out a notice to all departments, there would be a $10,000.00 increase in budget for every percent the accidents were brought down.

Next year the accidents were the same. They now decided this was unacceptable, it was time to get serious. They hired a consultant who studied all of the accidents at the intersection. The consultant found that out of ten accidents, three involved speeders and thatís thirty-percent. They lowered the speed limit on the roads, but next year the accident percentage still remained thirty-percent higher than average.

Totally frustrated they held a meeting. After discussing everything they had tried it was decided the problem could not be solved. A janitor was cleaning in the rear of the room and raised his hand. He asked, "Why donít you cut down that tree that blocks the stop sign?"


1.) A numerical goal without a method is useless. With a proper method, it is unnecessary. Study and planning (PDSA) solves problems, not goals.

2.) Incentives attached to goals, without the needed changes will not be effective. An unsafe design is as unsafe with or without the incentive. Incentives generally start with the assumption that people are the problem and a reward will change them.

3.) Looking at data without understanding the system is ineffective. The consultant could have studied the system as a whole. It may have been learned, thirty-percent of all vehicles going through all intersections were speeding. Since they represent thirty-percent of the total, statistically they will also represent thirty-percent of the accidents. Misuse of figures to justify a theory is ineffective.

4.) This intersection is a stable system that produces accidents at a rate 30% higher than desired. To change the rate, the system [intersection] must be changed.

5.) Many times answers come from unexpected sources. Many times answers come from outside of the system.

6.) You can substitute the words repair shop for intersection, low production, comebacks, low profit or whatever for accidents. Now you need only figure out which "tree" to remove.

View Comments (0)