What caused Sundays Metro-North crash, which killed four New Yorkers and critically injured 11? Its natural for people to look at two potential culprits—human error and equipment failure—and pick one of the two. Mondays late report that the train was going 82 miles an hour in a 70 mph zone as it entered a 30 mph zone has focused attention on the trains operator. But humans and infrastructure go together. After all, human beings make the infrastructure—and well-designed and maintained infrastructure can, in turn, keep all-too-human mistakes or malfeasance from resulting in catastrophes.
The Thanksgiving weekend crash is getting lots of attention, as it should. Trains dont just crash, just as planes dont just fall from the sky. The crash fatalities were Metro-Norths first in the commuter railroads 30-year history as part of the state-run Metropolitan Transportation Authority. Despite the accident, youre still 10 to 11 times safer on the train than you are in a car or small truck. The biggest risk many suburban New Yorkers will take this week is driving to work or to a different train station because regular service is curtailed.
But a series of early-warning mishaps was already hurting Metro-Norths stellar record of moving 84 million people safely and reliably every year. In May, a pre-rush-hour afternoon train out of Grand Central to New Haven derailed, causing another train to crash into it. Nine passengers required hospitalization, two for prolonged stays. Also in May, a train struck and killed a track worker. Two months later, a private garbage train derailed near the site of this weeks crash. And in September, a power failure knocked out service from Connecticut for a week.
Each of these incidents shows how much can go wrong. There are the rails themselves. Its easy to think of rails as inert objects—and of defective, broken, or badly installed rails as an infrastructure failure, not a human failure. But to keep rails in safe condition is a massive task for people. Last year, Metro-North had 1,637 maintenance of way workers inspecting and repairing rails—more than 27 percent of its workforce. Thats 125 more inspectors than in 2008, meaning that direct staff cuts in the wake of budget austerity likely werent to blame in any of this years wrecks.
But despite this manpower, Metro-North isnt doing a good enough job of inspecting rails and replacing old ones. Were behind in several areas of our [maintenance] programs, chief engineer Robert Puciloski told investigators looking into the May crash at a federal hearing last month. I cant give you an answer as to how we got so far behind. One pressure is that workers must inspect tracks quickly—too quickly—to allow trains to pass. Last month, Metro-North chief Howard Permut said that the railroad is improving drainage and replacing some concrete ties with wood ties, which may perform better. The railroad is also looking at better technology, with new ground-penetrating radar to identify and repair areas of track that could eventually cause problems. But he noted that a significant amount of work remains to be done, hardly a comforting thought.
The track-inspection failure highlights another problem: management. Metro-North has slashed administrative headcount over the past half-decade, from 726 people to 469. Some of these cuts have been for the good. As the MTA has merged different offices and phone banks, for example, Metro-North customer-service ranks have fallen from 243 people to 39. Thats fine, if people can get their information and tickets from computers and kiosks. At the same time, safety-administration worker ranks have barely fallen—from 18 to 17, theoretically showing that the railroad has prioritized safety as it has made its budget cuts.
Yet Metro-North employees and retirees have long warned that a slimmer core of middle management—plus a misguided focus on on-time performance at all other costs—has harmed safety. We have had an exodus of retirees taking with them the skills and experience needed to run such a dangerous industry, one former worker told me after I wrote a piece cataloging bad-omen derailments in early summer. The May death of track foreman Robert Luden after a trainee opened a track that was supposed to be closed was a clear indictment of Metro-North supervision. And earlier this year, Metro-North scrambled to hire 11 new managers just to improve safety, implying that it hadnt had enough managers before (or that it didnt have the right managers but cant fire the wrong ones).
Another key issue is long-term investment, which can be as simple as keeping leaves off the tracks or as complicated as building an automated train-to-train communication system. Take the issue of autumn leaves. Fallen leaves can create a gelatinous substance on tracks that causes wheels to slip or slide. Technology can automatically detect such slippage, even stopping a train. But such investments cost money. For its current five-year capital plan, Metro-North originally planned to spend $334.6 million on track and other structural upgrades over five years. Instead, its spending $301 million. On signal and other communication upgrades, it was going to spend $278 million; now, its spending $273.4 million. As debt costs for past capital investments as well as employee benefits inexorably rise, the MTA has had to cut back on new capital spending and likely will make deeper cuts in the future.
Its not just the amount of capital money thats spent, though, but how its spent. Just last month, for example, the MTA approved a $210.5 million long-term contract to install positive train control, a system that the federal government now requires in the wake of an earlier crash in California. Positive train control allows trains to detect excessive speed as well as the presence of other trains. Such a system may have prevented Sundays crash by detecting and halting dangerous speeding, NTSB official Earl Weener said Tuesday afternoon. But, as the MTA warned, positive train control is complicated. It involves supervising contractors who must build five separate electronic and detection systems as well as integration with Amtrak, whose trains share MTA tracks (think CityTime, but harder). Its so complicated and expensive, in fact, that the MTA is asking for an extension on its completion deadline from 2015 to 2018.
And positive train control is not a panacea, anyway. As the MTA also warns, it will consume at least $1 million annually in maintenance costs. (That figure is likely a placeholder; its hard to see how the MTA would maintain such a complex system for under eight figures annually.) A poorly maintained positive train-control system—maintained, remember, by people—is dangerous, too. Nearly half a decade ago, two trains on Washingtons Red Line Metro crashed, killing nine, partly because of a malfunctioning automatic system. Infrastructure decisions can make a difference after a crash as well. Did poor train-carriage decisions cause people to crash unnecessarily through sharp metal or glass?
All of this bears keeping in mind as observers—including Governor Andrew Cuomo—point to speed as a cause of the crash. Its quite possible, of course, that train operator William Rockefeller was recklessly speeding, as a Spanish train operator was earlier this year in a crash that killed 79. If he was, though, who or what was monitoring historic speeds in the area, for example, to make sure operators werent going too fast? Even 18-wheelers have speed governors, so that the boss can know, and even control, truck drivers speeds. Why wasnt the conductor on boardwho is officially in charge of the trainable to notice the high speed and intervene? Why didnt other, old-fashioned on-board controls—including what a former Metro-North worker described as a deafening whistle to warn of excessive speed followed by automatic braking—do their job? I do not know how this engineer could possibly have been going 82 mph, the Metro-North veteran told me. That is faster than any of the speeds on the railroad. . . . I do not know how he could go that fast with the speed restrictions built into the equipment. . . . It does not make mechanical or psychological sense.
As people pile blame on Rockefeller, remember, too, that it is in Cuomos interest, as well as in the MTAs, to place blame on one person, thus averting public attention from all the fail-safes that are supposed to mitigate the impact of that one persons mistakes. That one human being could possibly cause such a catastrophe is ultimately the failure of other human beings. And that means that Metro-Norths earlier failures, pointing as they did both to management and infrastructure dysfunction, were signs of worse to come. Remember, the pilot is not supposed to be able to crash the plane by pressing one wrong button.