Saturday, July 13, 2013

On selective endpoints

Shorter Andrew Coyne:
If you ignore the actual recent rail disaster that blew up a town, there's little apparent risk of rail disasters. So let's keep assuming the likelihood of future disasters is trivial.
Now, some observers might ask how consistent a particular event is with a set of assumptions about the probability of that event occurring, and at least question the "everything's fine, nothing to see here" thesis in light of an actual disaster resulting from the blatant neglect of what they themselves observed to be obvious risk factors. But apparently not Coyne.

And some of us might also point out that the purpose of a public safety regime - particularly one governing the transport of products with the ability to cause massive destruction - should be to establish policies which keep us a reasonable distance away from the borderline between safe operation and disaster, rather than to test the limits of what we can afford to abandon given the human consequences of getting the answer wrong. (In effect, this goes to the question of what "tests" we're willing to have carried out on an unsuspecting public in gathering evidence as to what is or isn't safe.) But once again, apparently not Coyne.

Fortunately, plenty of others are working not only on figuring out what happened in Lac-M├ęgantic, but also asking more fundamental questions as to what we expect from our regulatory authorities before the "actuality" of similar tragedies materializes. And if Coyne ends up alone alongside the Cons' know-nothing camp, then so be it.

No comments:

Post a Comment