I’m part way through reading Disastrous Decisions, Andrew Hopkins’ recent book on the Deepwater Horizon oil well disaster in the Gulf of Mexico. One chapter looks at why the people responsible for deciding that the well had been properly cemented did so in the face of considerable evidence to the contrary. There are multiple reasons but prominent among them are consensus decision making and groupthink, which occurred with the best of intentions but lead to decisions that were incorrect and ultimately tragic. I read this with increasing discomfort because group decision making is seen as central to a safety management study workshop.
Hopkins does not seem to be suggesting complete abandonment of participatory decision making but mentions other approaches that retain the advantages of participation while avoiding the worst features of groupthink:
- Decisions are made by a single responsible person after consulting with all relevant experts, OR
- Decisions are made by a group but the group includes an outside expert who is independent of the in-house team, OR
- Decisions are made by a group but one member of the group is appointed devil’s advocate to openly challenge the group’s thinking
In principle a pipeline safety management study has to be approved by the pipeline licensee so some might argue that the first alternative above is already in place. In practice however this is most unlikely to be effective unless the licensee’s responsible officer participates in the workshop; that hasn’t happened yet in the several dozen workshops I have been in.
For the second option, I guess that I often fill the role of independent outside expert myself when I run a workshop for a client, except that if I lead the discussion (which is very often the case) I could conceivably be part of the groupthink problem rather than a solution. I try hard to avoid railroading a workshop but there is a fine line between trying to provide guidance and inadvertently providing firm direction.
For that reason I quite like the devil’s advocate concept. I will be thinking about how it might be incorporated in future SMS workshops.