On glibertarianism

One of the more obvious points of convergence in political thought over the past 70-80 years is the greater appreciation of systemic complexity - the recognition that different decisions by many types of actors may collide in unpredictable ways, with the results potentially far outweighing the perceived impact of any single action and reaching people who may have had no idea their interests were affected.

That recognition in turn gives rise to two broad public policy responses: to seek to have a government equipped to identify the risks posed by factors which no other party has the means or will to address, or to give up and leave ourselves at the mercy of forces which interact in ways which we don't even try to anticipate.

Suffice it to say that Andrew Coyne's response to the Lac-Mégantic rail explosion is staking out some rather extreme territory in the latter camp.

Here's Coyne trying to isolate the specific circumstances of the explosion in order to treat the issue as one which doesn't call for an immediate response:

Consider what a singular convergence of events was required to bring it about. A highly flammable cargo; an unattended train; parked on a hill; on the main track, not a siding; above a town; far enough from town to build up great speed; and, as a final piece, that fatal bend in the track as it entered town. If any one of those is not present, no disaster and no deaths. But even if all are, you still need two more: the failure (so it seems) of the air brakes; and the failure (so it is alleged) to lock the hand brakes.

Now, may be true that Coyne's list of factors is indeed extremely unusual. But I'm fairly sure the right answer to "how many towns could see the same type of disaster?" is "we don't know, and should probably find out" - not "one - and since it's already been blown to smithereens, let's not worry our pretty little heads about it."

And indeed, a closer look at those same factors reveals some rather important decisions not to consider possible public safety issues:

Consider what a singular convergence of events was required to bring it about. A highly flammable cargo whose transportation has never been subject to a specific regulatory evaluation; an unattended train which is permitted by regulation; parked on a hill which should presumably be known to the rail operator and regulators alike; on the main track, not a siding which should presumably be known to the rail operator and regulators alike; above a town which should presumably be known to the rail operator and regulators alike; far enough from town to build up great speed which should presumably be known to the rail operator and regulators alike; and, as a final piece, that fatal bend in the track as it entered town which had been pointed out by concerned residents. If any one of those is not present, no disaster and no deaths. But even if all are, you still need two more: the failure (so it seems) of the air brakes; and the failure (so it is alleged) to lock the hand brakes.

Again, this brings us back to the core question of how to deal with complexity. Should we simply ignore risks in the name of minimizing our own costs, presuming that somebody else (and in this case, a private operator devoted to shipping the most product for the least cost) will bother to identify what needs to be done in order to avoid public tragedies? Because the evidence from Lac-Mégantic (and an ongoing series of spills, derailments, explosions and other disasters) doesn't exactly support the "let somebody else do it" argument.

Moreover, it seems equally obvious that other "convergences of events" might come about elsewhere. And that's precisely why we need a regulatory structure equipped to evaluate the distinct threats facing all sorts of communities and ensure that they're being avoided to the extent possible - rather than one based on allowing private operators to self-report on standardized checklists as the primary precondition for carrying out dangerous activities.

But sadly, we have a government more interested in slashing public services than making them work - which is an issue we should be discussing regardless of whether a particular regulatory tweak would have been necessary or sufficient on its own within the complex array of factors which might have stopped the Lac-Mégantic tragedy. And yes, the aftermath of a preventable disaster is probably the best possible time to highlight the fact that we should be making some effort to prevent disasters.

Speaking of which, I haven't even reached the most galling part of Coyne's take:

There are things we could do, without a doubt, that would preclude another Lac-Mégantic altogether. We could make the cars out of titanium, or reroute the lines around towns, or take out the bends. But are the costs, potentially high, worth the risk: vanishingly small? We could ban carrying oil by train, but other methods, as I’ve mentioned, have their own risks — and what of the vast number of other hazardous materials that also travel by rail?

Put in its most charitable light, one might read this passage as referring solely to probabilities - to the exclusion of any talk about the cost imposed on innocent parties should the possibility materialize. But surely any rational calculation of risk includes some consideration of both. (Paging Dan Gardner on this one.)

And I'm far from willing to consider this...

...to be a vanishingly small cost incurred in the fulfillment of the apparent highest possible good of ripping carbon from the ground and burning it.