Railroad travel was astonishingly dangerous in its early stages in the US. Trains derailed and collided; passengers died trying to catch a train or trying to jump off; children played on the tracks; strap iron rails detached from their wooden base, poked through car floors, and impaled people; boilers and cargos exploded. Until the later part of the last century, it was more dangerous to be a railroad passenger in the United States than a railroad worker in Britain, by about a factor of three.


Author Mark Aldrich lays this all out in Death Rode The Rails. Aldrich is primarily a economic historian, and I started the book with a little trepidation, because I had already read a railroad economics book (The Most Valuable Asset of the Reich) that was just ghastly. This one, however, is not that bad. Aldrich is very evenhanded for placing blame. Although it's not something you want to read on the beach, it pleasant to dip into periodically. Some examples, in no particular order:


The early American railroad system served a very low density population compared to Europe; it didn't make sense to do the elaborate track ballasting and bridge construction that European railroads used. Therefore, a lot of rails were laid without any ballast at all, and a lot of bridges were built of timber instead of stone. Ironically, while visiting European passengers were utterly appalled by American railroads, visiting European engineers praised them as being an excellent adaptation to circumstances.


European railroads had fences between the station platform and the tracks; passengers entered through a gate and boarded the train from the side. It made no sense to run after a train because there was no way to board even if you could catch up to it. American stations had no separation between platform and train, and you boarded at the end of the cars - running after a train (and falling into the wheels) were therefore feasible options.


European railroads adopted block signals very early; American roads stayed with train orders much longer. When American roads began introducing blocks, they were usually "permissive" blocks; after coming to a stop the train could proceed into a stop-aspect block as long as the engine driver had the train "under control". There's a flip side to this; although it isn't mentioned in this book, government-mandated block signalling in Britain (according to Railroad Signalling) slowed the adoption of other safety features.


Compensatory behavior was very evident. When the railroads introduced air brakes, improved bearings, stronger rails, and stronger trucks, worker safety didn't improve much. Instead, railroads began running heavier and faster trains.


The biggest improvement in worker safety was in institutional changes, rather than technological ones. Management began holding safety meetings and disciplining workers who were caught ignoring safety rules (this change had to be communicated to lower level supervisors, who previously we judged based on their production performance rather than safety performance).


Labor unions, on the other hand, generally pressed for technological changes rather than work rule changes, and often disguised full employment measures as "safety" rules - for example, requiring multiple brakemen on trains equipped with air brakes, and demanding a fireman on Diesels.
Aldrich isn't all on management's side, however; he criticizes the railroads for not involving the unions in safety programs early on, which led to hostility and problems later.


The "car interchange" procedure in the US, where one line's freight cars would be swapped with another, caused a number of problems. Railroads were very reluctant to repair competitor's cars for them. In fact, some roads would strip good parts off a competitor's car and replace them with cheap ones before returning it.


There are some subtle interactions that aren't immediately obvious. For example, pension plans apparently improved safety, because it gave workers incentive to remain with a company and gain more experience. One of the safest times for workers was the Great Depression, because many workers stayed on the job for years. (The Depression was, of course, a bad time for trespasser accidents, because a lot of people tried to hitch rides on freights).


There are also subtle interactions between the railroads and automobiles. The introduction of the automobile reduced trespassing accidents at the same time grade crossing accidents increased. Trespassers walking the rails were often trying to get from one town to another, and a railroad ROW could be much more convenient and direct than dubious horse-and-buggy roads.


The US Government played a mixed role. Some government mandates seemed to be beneficial, while others were counterproductive. The government (as is still true today and probably always will be) tended to respond to perceived risks rather than actual ones, concentrating on rules intended to prevent spectacular derailments and collisions, rather than day-to-day worker and passenger deaths, even though the later took many more lives in the long run. A case in point is the forced introduction of "automatic train control", which were devices that would automatically stop a train if it entered a block with a "stop" aspect signal. ATC was complicated, never worked well, and didn't do much to improve safety (because most accidents were one-at-a-time rather than spectacular collisions). The railroads eventually successfully petitioned to have the ATC requirement removed.


The railroads developed the hazardous material transport standards now administered by the USDOT. There were some interesting accidents before that; my favorite was a dynamite-loaded boxcar that began leaking nitroglycerin through a hole in the floor onto the track. Yards crews were amused by the popping and banging "fireworks" noises the car made every time it was moved. Eventually it stayed in one spot in the yard long enough to accumulate a substantial nitroglycerin pool, and when it was finally moved again the result wasn't amusing at all.


There are some excellent statistical tables. Unlike a lot of authors incorporating statistics, Aldrich is very careful to warn of "apples and oranges" problems, where different states and different railroads had different standards for reporting accidents.


There are some minor flaws; although Aldrich is a reasonably engaging writer you can only print so many accident statistics before things began to dull. Also, I would like to see a diagram showing exactly how a Westinghouse Type K air brake works. It's explained several times in the text but I still don't understand it.


Fairly expensive if new off the shelf but I found my copy at a remainder outlet; as said, not light reading but handy to have around for reference. ( )

referenced in Sebastian Junger's "Freedom"