Death Rode the Rails, Indeed

The prospect of dying in a railroad accident is not something that too many Americans worry about these days.  But it was not ever thus.  In an excellent but little-known book entitled Death Rode the Rails:  American Railroad Accidents and Safety 1828-1965, economic historian Mark Aldrich reveals that in the earliest days of rail travel in the 1840s, passengers were sometimes surprised to see a thin strip of iron thrusting up through the floor of the carriage, threatening to impale them like bits of beef on a barbecue skewer.  Called "snakeheads" by the antebellum press, relatively few people were killed by these accidents, which occurred because some of the earliest railroads used a thin iron strap fastened to a wooden rail to save money, instead of a solid iron rail, and the strap would sometimes come loose from the wood, snaking its way up into the cars.  But the combination of surprise and powerlessness to avoid the accident made it particularly horrifying, and the novelty of rail travel was tarnished in the public mind by this vivid addition it made to the list of ways one could depart this earth.

While Aldrich has plenty of stories about the different ways that passengers, railroad employees, and trespassers on railroad property were injured and killed, his emphasis is on the economics of railroad safety and how economic considerations played a vital role.  He points out that even after wood-and-strap-iron rails were replaced with all-metal rails and many other safety improvements were made, traveling by U. S. rail in 1907 was still 110 times as dangerous as flying in a modern (2006) airliner.  Still, 22 fatalities per billion passenger miles did not mean that you were taking your life in your hands every time you climbed onto a train.

From the railroad companies' point of view, safety was an expense, and like every other expense, they wanted it to pay a return on investment.  Railroads were virtually unregulated by the federal government until the establishment of the Interstate Commerce Commission (ICC) in 1887, and for many years the ICC restricted itself to setting freight rates for interstate commerce.  Some safety ideas, such as the "block signal" system of controlling train movements, rather than sending out paper orders and hoping everyone would synchronize their watches and keep to the schedule, not only reduced accidents but increased traffic flow, leading to greater utilization of existing plant and higher profits.  The railroads liked this kind of safety measure.

On the other hand, in 1922 the ICC ordered all carriers (rail lines) with revenues over $25 million to install automatic train control on at least one passenger line.  The idea of automatic train control, which dates back to the 1800s, is that instead of relying on the engine driver to see a visual block signal and stop the train, the automatic system would directly receive the signal's command and apply the brakes.  The rail companies reluctantly complied, and by 1930 had spent $26 million to install the system on over 15,000 miles of track. 

But as Aldrich shows, automatic train control made essentially no difference in the rail safety record, did not improve productivity, and cost a great deal of money.  During the Great Depression, many carriers asked for and received permission to cut back or remove automatic train control, and the ICC relented.  However, the same technology turned out to be useful for activating signals in the driver's cab (so-called "cab signals"), which have since become a standard safety feature of great help in fog or rain where visibility of the track-side signals is obscured. 

I was unaware of all this when I blogged a few years ago about a "cornfield meet" (head-on collision) between two freight trains in Texas that killed three employees and did millions of dollars of damage.  At the time, the railroads were installing something called Positive Train Control (PTC), which is nothing but an updated electronic form of automatic train control.  So the idea has been around for more than a century, it turns out, and is just now being implemented.  But as Aldrich points out, the accidents in which PTC would have made a difference are a small percentage of all mishaps.

While Aldrich makes a great case that economics was a huge factor in railroad safety, he gives less emphasis to something that continues to drive debates about all kinds of transportation safety today:  public perception.  He does point out that the average citizen has an exaggerated horror of types of death that are grisly and out of one's control, such as the snakehead accidents.  All the statistics in the world will not comfort the lizard part of one's brain that is primally terrified by the prospect of a fiery or gory death inside some machine that you cannot influence.  But other factors, such as speed and convenience, can overcome such fears.  For example, early automobile travel (say around 1920 to 1940) was demonstrably many times as dangerous as rail travel, yet the rail lines lost most of their short-range passenger business to the automobile in that period.  Ah, but the driver of a car has at least the illusion of control, thinking that while accidents may happen to other drivers, his superior skills will enable him to avoid a crash.  Well, maybe, but the statistics said otherwise.

As you would expect, engineers come in for starring roles in Aldrich's saga.  The technical press, including editors of such publications as Railway Age, brought constructive criticism to egregious safety problems and coordinated cooperation among carriers, government institutions, and private and university researchers to bring about notable improvements in safety systems, devices, and training.  This included issues such as the quality of bridge construction.  Early U. S. railroad bridges were built with the "link-and-pin" method, and the failure of even one joint in the structure would make the whole thing fall down, which it often did.  Complex failure modes in steel rails baffled engineers and scientists for decades until a concerted effort involving inventor Elmer Sperry's electrical track inspection system and advances in metallurgy discovered how to prevent them. 

An old friend of mine summed up the goal of engineering ethics with the two-word phrase, "No headlines."  While U. S. railroads are doing pretty well today by that measure, it is the end result of many decades of improvements and safety efforts.  And Mark Aldrich has given us that history in a rewarding and highly readable volume.

Sources:  Death Rode the Rails (Johns Hopkins Univ. Press, 2006), by Mark Aldrich, is the source for most of my material.  I also drew on the following website for additional details about "snakeheads":  https://aaronwmarrs.com/blog/2012/02/snakeheads-on-antebellum-railroads.html.  My blog about the head-on collision in Texas is at https://engineeringethicsblog.blogspot.com/2018/08/some-answers-about-panhandle-cornfield.html.

Will We Miss Net Neutrality?

On Thursday, Dec. 14, the U. S. Federal Communications Commission voted 3-2 in favor of repealing the Obama-era "net neutrality" rules that have been in effect since 2015.  Like so many things lately, net neutrality has become a partisan issue, and the vote went along party lines, the three Republican appointees on the Commission voting in favor of repeal and the two Democrats opposing it.  Polls show that the idea of net neutrality is popular, with as many as 80% of those asked being in favor of it.  But the pollsters generally didn't ask respondents to define net neutrality, or to say why they favored it.  Amid the protests and shrill voices raised on both sides of the issue, it's hard to get a grasp on what exactly is at stake, and what the pros and cons are.  A little history may help in this regard.

Among other things, most modern governments are expected to protect the weak against the strong.  This is an elementary aspect of justice.  In the late 1800s, during the rapid expansion of another kind of network—the railroad network—the public became aroused over perceived abuses that the railroads were practicing.  Farmers discovered that the railroads were manipulating shipping charges to curry favor with certain interest groups, and handing out free passenger passes to influential politicians.  The problems were so pervasive that the first free-standing administrative commission in the executive branch of the federal government was established to ride herd on the railroads:  the Interstate Commerce Commission, or ICC.

The ICC established rules for what became known as "common carriers"—enterprises that were so essential to the public that regulation by government was regarded as necessary.  The idea of a common carrier spread to other systems such as bus lines, airlines, and public utilities like electric and water systems.  In exchange for close regulation by the government, the business being regulated was allowed to make a reasonable profit.  Some industries eventually came around to welcoming common-carrier status, because they found that manipulating the government's rules in their favor wasn't that hard and it stabilized their business models. 

In 2003, a Columbia University professor named Tim Wu coined the phrase "net neutrality" to extend the common-carrier idea to the internet, which was not regulated in any meaningful way at the time.  In the case of the internet, the potential for the kind of abuse that the railroads got into trouble for is always there.  And there have been some incidents prior to the 2015 adoption of formal net-neutrality rules that give advocates of net neutrality some credibility.  According to the Wikipedia article on net neutrality, the internet service provider (ISP) Comcast took measures to throw digital roadblocks in the way of the troublesome service BitTorrent, which was using up a lot of bandwidth at the time, and the FCC has fined AT&T for similar misbehavior.

But the net neutrality rules that the FCC has now pledged to abandon may go too far in the other direction.  According to ISPs, the rules left them with limited flexibility for expansion and the offering of new services.  Treating everybody the same on the internet is a fine idea in principle, but working out the details can get complicated, and there are genuine judgment calls involved in an ISP's decisions of how to allocate limited fiber-optic and especially wireless bandwidth to best serve the incredible variety of customers, and websites that customers want to visit. 

We have seen how the content providers themselves (e. g. Facebook) have done things that go against some principles of net neutrality, such as the idea of no censorship.  Both for legal and moral reasons, Facebook polices itself and removes posts it deems to be unsuitable for various reasons.  But it's not an ISP that's doing this, it's Facebook. 

The ISPs, as ISPs, do not have the resources (or I suspect, the inclination) to do a lot of fine-grain discrimination, which is probably the kind of thing that many people who favor net neutrality are worried about.  Basically, the ISPs don't have time to pick through the floods of data that they must ship around every microsecond.  The most they can do in a typical situation is to note sites and services that produce unusually demanding traffic patterns.  And I think the most that they are hoping for in the repeal of net neutrality is to gain some freedom more efficiently to allocate their bandwidth in order to serve the most customers with the fewest additional resources of hardware and software. 

Maybe that is a Pollyanna-ish and naive view of ISPs, but it's hard for me to imagine that some of the more dire consequences foretold by the proponents of net neutrality will result from its abandonment:  widespread censorship, the inability of small-scale websites and enterprises to compete with larger ones based on something the ISP is doing, and so on.  One concern, transparency, is largely being taken care of by the internet itself.  Tricks like artificially degrading services are quickly detected and exposed by users, and it's easy for protesters to gather a digital lynch mob with torches and clubs and go after the bad guys.  Whether the bad guys mend their ways is another question, but my point is that if an ISP tries anything unpopular, they will be called out for it.  And this is an important self-regulating aspect of the internet that we may not appreciate as much as we should.

So my own answer is, no, I don't think we'll miss what we've had for only the last two years anyway, in terms of the Obama-era net neutrality regulations.  Even critics of the FCC decision admit that nothing is going to change right away, as the Commission has to come up with alternative rules and perhaps turn over some aspects of its work with the internet to the Federal Trade Commission. 

The internet is a modern necessity, not much less essential than electric power, and it is appropriate for governments to make sure that whoever qualifies as "weak" with regard to it is protected against unfair and unjust depredations by ISPs, or anybody else for that matter.  But even in the bad old days before government regulations were in place, abuses were fairly rare.  And it looks like the commercial instinct of self-preservation will keep ISPs from doing anything really dastardly, now that net neutrality rules are going away. 

Sources:  I referred to reports on the FCC vote to repeal net neutrality carried by AOL.com at https://www.aol.com/article/news/2017/12/14/fcc-commissioner-closes-statement-on-net-neutrality-vote-with-a-warning/23307786/ and https://www.aol.com/article/news/2017/12/14/federal-communications-commission-votes-to-repeal-net-neutrality-rules/23307670/.  I also referred to the Wikipedia articles on net neutrality and the Interstate Commerce Commission.