4.3 Root causes: programs, not behaviors

1 What a root cause is

The guide defines root causes as “the underlying reasons why unsafe conditions exist or if a procedure or safety rule was not followed in a workplace”, and adds that they “generally reflect management, design, planning, organizational or operational failings” — a damaged guard that had not been repaired; failure to use the guard routinely overlooked by supervisors to keep production moving.

And it gives the test for whether you have reached one: “Eliminating the immediate causes is like cutting weeds, while eliminating the root causes is equivalent to pulling out the roots so that the weed cannot grow back.” If the answer you have written would let the same sequence happen next month with a different person in it, you have cut a weed.

2 Why blame stops the investigation

The guide gives two reasons and both are worth stating out loud in a meeting where somebody is arguing for discipline.

The first is logical: “If an investigation is focused on finding fault, it will always stop short of discovering the root causes, because it will stop at the initial incident without discovering their underlying causes.” Fault is an answer, and an answer ends the inquiry.

The second is social: “if an investigation is understood to be a search for 'someone to blame,' both management and labor will be reluctant to participate in an open and forthright manner. Workers will be afraid of retaliation and management will be concerned about recognizing system flaws because of potential legal and financial liabilities.” An investigation nobody talks to produces nothing, and everyone knows within a week which kind of investigation this company runs.

3 Asking why until it lands on a decision

The method in step 3 is persistent digging. The guide works an example: if the procedure was not followed, ask why. If the answer is that the employee was in a hurry to complete the task and the safety procedure slowed the work down, ask why the employee was in a hurry. Each answer becomes the next question, and the deeper the questions go, the closer you get to something a manager can change.

1. 1
   Start with what happened, stated as a fact.
2. 2
   Ask why that was possible — not who did it.
3. 3
   Take the answer as the next question.
4. 4
   Stop when the answer names a decision, a design, a schedule or a program — something the company owns.
5. 5
   If the last answer still describes what a person did, you have another why to ask.

Expect more than one root cause. The guide says so directly: root causes often involve multiple deficiencies, in areas such as workplace design, cultural and organizational factors, equipment maintenance, operating systems and procedures, staffing, supervision and training. A single-cause conclusion is usually an incomplete one.

4 The five families of contributing factor

The guide's own examples of the questions an investigator pursues fall into five groups, and running through them systematically is what stops an investigation from following only the most visible thread.

Appendix F of the guide extends this into thirty-three numbered questions covering procedures, environment, space and lighting, familiarity and deviation, tools and PPE, emergency procedures, equipment history, staffing levels, supervision, job review sessions and training. It is a long list and it is meant to be worked through rather than read.

5 The table saw example, worked

The guide's own illustration is an amputation on a table saw, and it shows the shape of a systems inquiry in four questions: was the machine adequately guarded, and if not, why not? Was the guard damaged or non-functional, and if so, why had it not been fixed? Did the guard design get in the way of the work? Had the employee been trained properly in the procedures to do the job safely?

Notice what the third question does. “The guard got in the way” is usually treated as an excuse. In a systems approach it is a finding about design: a guard that makes the job harder will come off, on every shift, in every company, and the fix is a different guard rather than a different attitude.

6 Writing the root cause statement

A usable statement has three properties. It describes a condition or a decision rather than a person. It explains why the condition existed and why it had not already been addressed — the guide is explicit that when a shortcoming is identified, it is important to ask why it existed and why it was not previously addressed. And it points at something a named person has the authority to change.