The Perception of Risk
The international maritime community is continually engaged in a process of improving safety at sea, within the bounds of technology, cost-benefit analysis, and the ability of said community to agree. Big changes in maritime safety are almost always in response to an accident, or an accumulation of accidents. Some have more leverage than others to open a “policy window” and push the stone of regulation a little farther up the hill of inertia and disparate interests. One of the more ambitious attempts in recent decades to foster a culture of safety at sea is the International Safety Management (ISM) Code. Its creation was largely driven by the sinking of the HeraldofFreeEnterprise in the English Channel in 1987. When the ferry left port with its enormous bow doors open, the North Sea washed in and the vessel immediately capsized taking 188 lives. The investigation found not only lax practices aboard, but also a dysfunctional shoreside management apparatus that placed sailing on time above safety. Among other things, the ISM Code codifies a process of continuous improvement in safety practices, which includes bridging the gap between shipboard and shoreside perspectives.
The ElFaro fell under the ISM Code and had an up-to-date Safety Management System in place. There is every reason to believe that the ship met all applicable standards at the time she sailed, though there is much we do not know. This much we do know: It is easier to codify safety equipment, licenses, hull thickness, renewal dates, and service intervals than an individual’s or an organization’s perception of risk. In general, we can assume that long experience leads to a more accurate perception of risk, and therefore more appropriate margins of safety. But occasionally it may not, especially if past experience leads us to overestimate our ability to control events, and underestimate the capacity for things to go wrong in the vicinity of a hurricane. Giving decision-makers the latitude to make only good decisions but never bad ones is a conundrum that no code has yet solved.