"Is it rational to minimize the expected utility?" - The Northern Sea Route case

The Swedish risk philosopher S.O. Hansson published in 1993 a paper with the title “The false promise of risk analysis” were he discussed five problems with risk analysis as it was used then (and now). The fifth problem was that it has to be rational to minimize the expected utility, which among other things means that the risks discussed has to be comparable. One example of risks that are not comparable is two risks with about the same expected value (probability times consequence) but with totally different levels of consequence, i.e. one very unlikely catastrophic event compared to a rather common event with a low level of consequences.

The Northern Sea Route case:

At the moment the ice along the Northern Sea Route, from Asia to Europe to the north of Russia, is at its season low and the amount of ships passing thru the route this year is expected to be ten times the number of ships that passed through the route only three years ago. The northern sea route takes weeks off the voyage which off course saves resources such as fuel. The route also at the moment is pirate free. Therefore the northern sea route is a win for everybody (except for the ones making money on ships taking other routes) the ship owners, the consumers but maybe especially the environment. This is true until there is accidents on the route were the cold climate makes the consequence of an accident so much severe. An oil spill in cold waters is very problematic, but also other consequences of an accident are worse because of cold water and scarcely populated areas.

If we would calculate the environmental risks for a ship with the traditional route thru the Suez Canal and compare them to the risks with the northern passage my GUESS is that the risk for the traditional route is higher, but the risk with the northern passage is potentially more catastrophic. Which is worse? I don’t know but the environmental organizations in media seem to think that the northern route is the worst alternative.

So here we have a good example of a case where it is hard to compare the expected risks, but also where it is difficult to obtain reasonable utility estimates which is Hansson’s third problem with risk analysis.

Reference: Hansson, S. O. (1993). "The false promise of risk analysis." Ratio-New Series 6(1): 16-26.