Congestion, population growth, and safety

Today, I'd like to take a look at a possible alternate way to view the frequency of railroad accidents.

I don’t want to fill the first half of this blog post with a list of recent railroad accidents. If we’re paying attention, we see reports of derailments, deaths, grade crossing collisions, and near-miss or no-miss collisions between trains every day. As I write this, there have been two such accidents in the past 24 hours, five in the past week, including at least one fatality.

Statistics can be manipulated to demonstrate just about anything to those who aren’t thinking critically about what is being demonstrated.

The latest manipulation says that railroad accidents are down. In reality, according to the FRA (United States Federal Railroad Admin), the number of accidents are flat when compared with last year but down over a 3-year period. Total fatalities are actually up over both comparison periods.

Non-fatal injuries are down over both periods, but this is only true if you look at all accidents involving railroads. When you look at non-fatal injuries involving only trains (excluding vehicles, premises, falling off a bridge, and the like) you see non-fatal injuries up a whopping 369 percent over the 3-year period.

And therein lies the crux of my argument. Railroads are only getting safer to the extent that congestion, population growth, and complexity of operation will allow.

There was once a time when a railroad accident rate would be dependent on the huge number of people employed by the railroads. The interactions with the general public were, of course, meaningful in terms of statistics, but insignificant when compared to the risk taken by railroaders on a daily basis.

This is no longer true. The railroads have pared down the fleet, pared down employment, modernized, safety trained, and in general become just about as safe as any industry out there with, perhaps, the exception of long haul trucking. An uptick in railroad employment is a statistical margin of error in an increase in accidents.

Railroads, like transportation businesses in general—with the possible exception of river and coastal waterway traffic—interact with the general public at all hours of the day and night. I’m not just talking about customers, but about people who are otherwise unlikely to interact with a railroad for any reason. I argue that, for any given statistical area, as population and population density increase, the probability of an accident also increases. The more people cross a railroad track in any given time, the more people will be hit by a train. The denser the population, the more likely someone will trespass to make a trip shorter.

The density argument can be extended to congestion, both railroad and other types of transportation. The slower a train gets (from congestion), the longer it occupies a particular part of space in any given period. The longer a train is present, the longer it is interacting with the public, and the more likely an accident will occur. Applied to non-public-interaction accident, like derailments, the congestion argument says that the longer a train or trains occupy a track for a given period, the less of that period can be devoted to inspection and maintenance.

Then there is the chaos of complexity. By Amtrak’s calculation, 72 percent of its miles traveled are over so-called “host” railroads. I liken this to visiting your neighbor with a precocious child. Theoretically, the average home should be safe for all children, but a neighbor who does not have child locks and keeps bleach under the sink, though not breaking any regulations, is endangering your child. Moving on one more neighbor down, this neighbor might have a window at ground level. You may have put Plexiglas in yours, but his is ordinary window glass and will cut your kid to ribbons if he or she runs into it.

In like manner, when Amtrak (or for that manner any passenger railroad agency that operates on non-owned track) moves from one “host” to another, things change, even if all regulations are followed. Dispatching training can be different, signaling may vary, workloads and priorities change. The more such changes to account for on any given trip, the more likely an error, and perhaps a resulting accident will occur. And, as Amtrak seeks to privatize routes, the likelihood of error will grow, not shrink.

What can be done? Little, as long as our society's trends promote the increased use of rail as a transportation mode. No more than traffic accidents will tend to go down in number as we put more automobiles on the roads. There’s a trade-off, and I think you can see it if you think about it. Increased use of passenger rail as opposed to private auto increase the safety of our world in other ways, like less auto pollution, for example; but fewer derailments may be out of the question.

©2017 – C. A. Turek – mistertrains@gmail.com

(Charles A. Turek is a writer and novelist based in Albuquerque, NM. After four decades working in areas of the insurance industry related to transportation, he now writes on all aspects of American railroading. Charles is a political conservative but believes in public funding of passenger rail as a part of the federal government’s constitutionally conservative obligation to provide for defense and public infrastructure so that private enterprise may flourish.)