Rail advocates have been quick to come forward both after last year’s Metrolink tragedy and yesterday’s DC Metro accident that commuting by rail is far safer than commuting by car.

These kinds of statistics aren’t very useful, as Lee Clarke points out in Worst Cases. He shows that while it is safer to travel by car when you measure per mile, it’s safer to travel by plane when you measure in per hour of exposure. I suspect that it is true rail commuting is measurably safer regardless what person-level measure of exposure you use.

What concerns me, however, isn’t the comparison: it’s what the Metro crash suggests about rail congestion (yes, there is such a thing, and DC is looking at it) and what that rail congestion means objectively for rail safety. I suspect that it is getting worse, and that that is itself a problem regardless of whatever risks are associated with cars.