Graded Homework

Good and Evil on the Rails

As a child Robert M. Sanchez counted
the cars on passing trains. One day when he was seven he ran toan idling
locomotive and the engineer took him into the wondrous machine, let him blow
the horn, and, unwittingly, set his course for life. As he grew up he often
visited nearby rail yards, never losing his fascination with trains. After high
school he drove Greyhound buses for a time and then found work with Union
Pacific on a maintenance crew. After several years he worked his way up,
fulfilling his dream of becoming an engineer. Soon Amtrak hired him. He and his
partner, a waiter, bought a home near Los Angeles. Neighbors de-scribed Sanchez
as relentlessly cheerful, buoyant, and passionate about trains. Yet trouble was
there too. He was caught shoplifting at Costco, pleaded guilty, and served 90
days in jail on weekends. He argued with his partner and suggested they
breakup. On February 14, 2003, his partner hung himself in their garage,
leaving a note that read: “Rob, Happy Valentine’s Day. I love you.” 1 Two years
later Sanchez became an engineer forMetro link, a commuter rail system crossing
six Southern California counties. Metro link carries about40,000 passengers a
day on a busy 388-mile track net-work shared with freight traffic. He loved his
job though he worked a tiring split shift. Soon he bought a modest suburban
house where he lived with four miniature greyhounds. Again, neighbors described
him as cheerful, spirited, and exhilarated by railroading, but some saw him as
a recluse who kept to him-self and avoided revealing his past. He abided with a
dirt yard that stood out in a neighborhood of tended landscapes. 2 Although
friends said Sanchez found joy in his work, there were a few difficulties. He
received five informal discipline letters for absences and failure to follow
rules. Twice he was counseled orally about use of his cell phone while on duty.
In July 2008 a suicidal man sidestepped a crossing arm and ran in front of the
train he was operating. Under Metro-link’s policy he took some days off before returning
to work, but, according to his family, he was forced togo back before his
emotional recovery was complete.

Friday, September 12, 2008

On this day, Robert Sanchez was up
before dawn. Here ported at 5:30 a.m. and worked four hours, rested four hours,
then returned to work in the afternoon. At 3:03 p.m. he took train 111, a
diesel-electric loco-motive and three passenger cars, on a commuter route out
of Union Station. After five stops he approached the Chatsworth station 33
miles northwest, passing a solid yellow light indicating he should be prepared
to stop at the next signal. He failed to radio the dispatcher and call it out
as required. It was a beautiful day there with clear skies, calm winds, anda
mild 73 degrees. After stopping for 57 seconds the train departed the station,
a random assembly of 225 souls with perhaps the most troubled in the lead. At
exactly4:20:07 p.m. Sanchez shifted the throttle from the idle position to
position 2 and released the train’s air brakes. As it moved, he pushed the
throttle to its maximum 8 position. Rapidly, the train increased speed to42
mph. At 4:20:20 he sounded the locomotive’s bell and horn for the Devonshire
Road crossing. 4 At 4:21:03 he received a short text message froma teenage rail
fan: “I would like that too. We al-ready need to meet 796. That would be best.”
This was about a plan for Sanchez to sneak him aboard the locomotive later that
day and let him take the controls for fun. At 4:21:23 Sanchez again activated
the bell and horn for the Chatsworth Street cross-ing. By 4:21:35 the train’s
speed was 54 mph and he moved the throttle back to position 4 and braked,
slowing it to 44 mph in preparation for a curve. At4:21:56 the train passed a
red signal light ahead of the curve. It was a command to stop. Sanchez failed
to radio in the signal and did not stop. At 4:22:01 Sanchez sent a text in
reply to the teenager: “yea . . . usually @ north Camarillo.” At4:22:02 the
train passed over a power switch turned to move a local freight train coming in
the opposite directionoff on a siding. The freight train was Union Pacific
LOF65-12 consisting of two locomotives and 17 cars. It entered the curve
eastbound at 41 mph as Sanchez came on at43 mph from the west. Closing at a
combined 84 mph, each locomotive became visible to the engineer in the other
only when they were 540 feet apart and four to five seconds from impact. In
that instant the Union Pacific engineer and the conductor, who was also in the
cab, saw the Metro link locomotive. The engineer hit an emergency brake and
started to run out the cab’s rear door. Seeing there was too little time he
“just stood there and watched it happen in disbelief.” 5 The conductor froze on
his feet, uttering an epithet. In the other locomotive, Sanchez did nothing
with the controls. At 4:22:23 the trains collided. The lead Union Pacific
locomotive crushed Sanchez before pushing the massive bulk of his locomotive
back 52 feet into the first coach. The compression killed 23 passengers.
Another person died in the second coach. A sheriff’s deputy described the
scene. “I saw locomotives en-gulfed in flames . . . and . . . I saw numerous people,
maybe a dozen, walking in various means, I don’t know, delusioned, like they
were zombies waking with various types of injuries with their hands outand
saying help . . .” 6 Rescue workers needed four hours to extricate all the
victims from wreckage. Hospitals took in 102 injured including the engineer and
conductor from the freight train.

THE INVESTIGATION

The National Transportation Safety
Board (NTSB)was called in. The NTSB is a small, independent federal agency
established by Congress in 1967 to investigate transportation accidents and
make safety recommendations. It did a detailed analysis of the collision,
interviewing witnesses, holding hearings, and examining physical evidence such
as the signal switch wiring and even fasteners on the track’s woodencrossties.
An autopsy found that Sanchez had adult-onset diabetes, high blood pressure,
and an enlarged heart. He met the clinical definition of obesity. And he wasHIV
positive. His use of prescription drugs kept these conditions under control.
The Union Pacific conductor’s blood and urine tested positive for marijuana
use, though this was not relevant to the cause of the accident. The
investigation also focused on management.Metro link is organized as a regional
association witha governing board of representatives from five Southern
California counties. It was formed in 1992 to improve mobility and reduce
traffic congestion in densely populated areas. Most of its operations are
outsourced. Sanchez was hired and supervised byConnex, the subsidiary of a
French corporation that ranMetro link’s trains under a contract worth about$25
million a year. Under the contract Metro link retained overall responsibility
for its operations. As one top Connexmanager noted, “We run the railroad the
way they want it run.” 7 However, much was delegated, including the supervision
of train crews. Connexconducted the “efficiency tests” required of every
railroad. 8 These tests are done by supervisors who observe trains, monitor
radio traffic, and analyze data from recorders in locomotives to check rules compliance.
For example, they use stopwatches to make sure engineers blow horns for 15
seconds before entering astreet crossing. They use radar guns to check train
speeds. They stop trains for surprise inspections.Connex supervisors performed
about 1,000 such tests monthly. During his three years with Metro linkSanchez
had only a few failures on them. In 2006,when a rule against cell phone use on
duty went into effect, a safety manager arranged for someone to callSanchez’
number, then stopped his train and boarded the locomotive. As they were talking,
Sanchez’ phone rang. The phone was not supposed to be in the operator’s
compartment or turned on, but it was stowed away in a bag and Sanchez said he
had forgottenabout it. The supervisor accepted this and simply counseled him
about the policy. No more calls weremade to his phone to test his compliance.
In 2007 he twice was cited for failing to call out a wayside signal. Engineers
are supposed to radio theMetro link operations center to acknowledge each
lighted signal they encounter. Still, his supervisor said Sanchez was
frequently tested on calling signals and his performance was “above average.” 9
Earlier that year Sanchez also got a written warning for neglecting to light a
marker at the end of his train. Andabout a month before the collision a
conductor sawhim using a cell phone as his train was ready to leavea station.
Sanchez told him he knew he should putthe phone away and did. The conductor
reported thisto their Connex supervisor, who spoke to Sanchezagain about the policy
and did two observations of him in the next two weeks. He was confident
thatSanchez understood the policy. However, the super-visor said it was hard to
enforce.

It’s almost impossible . . . [T]he engineers,
first of all, is going to have the door locked. You’ve got to un-lock the door
to get up on it. He’s probably going to hear you coming—he or she, and, you
know, itwould be almost impossible to surprise somebody,you know, to inspect it
. . . [O]f all the times I’ve gone up on a locomotive, I’ve never seen anybody
with a cell phone or talking on a cell phone.

In themselves, these incidents on Sanchez’
recordwere not damning. The Connex safety manager had asubjective faith in him.
“[He] was a competent engineer,” he told investigators, “[a]nd I felt comfortableputting
people with him.” 11 Several weeks before hisfinal shift Sanchez even got an
award for “safety andrules compliance.”

However, his behavior on the day of
the accidentshowed brazen deceit and disrespect for rules. Hefailed to call out
two signals. And Verizon Wirelessrecords showed he made four phone calls, sent
21 textmessages, and received 21 text messages while operating the train. It
was habitual behavior. On each of seven working days preceding the accident he
hadmade calls and sent and received between 30 and125 text messages while
operating trains. 12 Most of the texting was with teenage rail fans. Interviews
revealed he had once before let a teenager sneak onto run a locomotive. In its
accident report the NTSB stated the probable cause of the collision as Sanchez’
inattention to thered signal light because texting in violation of com-pany
rules distracted him. It made one new recommendation, that railroads put audio
and video devices in locomotive cabs to monitor train crews. It repeated a
previous recommendation for installing crash- and fire-protected cab voice
recorder similar to those in commercial airliners. And it noted that an
automatic system called positive train control would have intervened to prevent
the collision by taking control of the train when Sanchez failed to stop at the
red signal.

POSITIVE TRAIN CONTROL

Positive train control is an old
idea in railroading. It had been on the NTSB’s “Most Wanted List of Transportation
Safety Improvements” for 18 years at the time of the accident. Now, thanks to
Robert Sanchez, it would become a reality. Briefly explained, it is an
interconnected network of digital data and controls. It allows remote operators
to take control of trains from on-board engineers if necessary. It includes
these basic elements. Global positioning system receivers on trains to
continuously track movement. Computers on trains that record data and send
in-formation to displays in locomotive cabs about train position, speed,
length, and weight; route speed limits; actual and recommended throttle and
brake settings; sensor readings on cars; signal and switch settings; and more.
Wayside devices that monitor signals, switches, and track alignment, and can
detect overheated brakes, cracked wheels, rock slides, and other problems.
Wireless interfaces on throttle and brake controls that allow remote control.
Computers and displays in railroad operations centers that show the schedule,
position, speed, and control settings of each train in the network and allow remote
command of train and track functions. Modern train control is technically
complex, but the basic invention, electro-mechanical automatic braking, came
around 1900. In 1920 the Interstate Commerce Commission (ICC) ordered 49 railroads
to install it on passenger lines to reduce accidents and fatalities. Though
effective, the systems were very expensive to put in and maintain. When
interstate highways spread in the 1950s, rail traffic faced more competition
from trucking. Revenues fell, tracks were abandoned, railroads failed or
merged, and the ICC let companies discard the controls. After that, human error
regularly led to avoidable fatalities from train collisions,
overspeedderailments, and runaway locomotives in work zones. Periodic headline
accidents that killed passengers led to regular calls for reinstating automatic
controls. However, little was done because the rail-roads argued it was
unaffordable.

CONGREES ACT

When the National Transportation
Safety Board placed positive train controls on its “Most Wanted “list in 1990
it revived the issue. Congress considered action, but retreated when the
Federal Railroad Ad-ministration (FRA) did a study showing that the cost of
controls far outweighed safety benefits. 14 The FRAis part of the Department of
Transportation. As an executive branch agency its administrator is nominated

by the president and approved by the
Senate and, when appointed, reports to the Secretary of Transportation.
Congress created the agency in 1966 to regulate railroad safety. It also
administers federal programs that support railroads and promote passenger
service, giving it close ties with the industry it regulates. Most of its 900
employees have worked for railroads.

After the early 1990s there were
short bouts of Congressional interest in train controls after major rail
accidents. In 2003 Congress asked the FRA for an updated benefit–cost study. It
showed that the costs still far outweighed safety benefits. 15 In 2005 the
agency issued a rule to encourage voluntary use of train controls. 16 Lacking a
mandate, railroads in-stalled automatic systems on only about 4,000 track miles,
most in the Northeast.

A few legislators remained
interested in train controls. When the Metro link crash occurred, there were
two moribund bills in Congress, a House bill requiring controls on several
high-risk routes and a Senate bill seeking only further study. Neither was headed
to passage because of opposition from railroad lobbyists.

The Metro link fatalities mobilized
California’s two Democratic senators, Dianne Feinstein and Barbara Boxer, who
zoomed in like superheroes on a mission. Within a week they introduced an
amendment to the House bill, which had already passed, ordering rail-roads to
install positive train control. In remarks on the Senate floor, Senator
Feinstein grew irate and accused the railroads of “criminal negligence.”

The accident happened because of a resistance
in the railroad community in America to utilizing existing technology to
produce a fail-safe control of trains . . .Over the years the railroads
resisted, saying these systems are too expensive. Well, how expensive is the
loss of human life? The cost of any system doesn’t come close to the cost of
the lives that were lost this past Friday.

A week later she and Senator Boxer
invited Joseph H. Boardman, administrator of the FRA, to public hearing. Senator
Feinstein opened the hearing by saying she was upset with “lobbying behind the
scenes to prevent an early date” for installation of train controls. Boardman
explained to the two senators why “progress has not been faster,” namely
because of “limited availability of needed radio spectrum,” concerns about “interoperability,
“and “braking algorithms that need refinement.” 18 These technicalities must
have sounded likeexcuses to Senator Boxer and they drew a sharprebuke.

What powers do you have? What’s your
job? You’re sitting there saying you can’t tell them to do any-thing? . . . You
have the power, you don’t want to do it, you’d rather work for the railroads.

After the hearing Senator Feinstein called the
FRA “an old boys club.” “I think they sit down and talk to the railroads,” she
said. “I think they do what the railroads want.” 20 In floor remarks she tried to
stir her Senate colleagues to action with a moral argument.

When we know there is global
positioning that can be in place to shut down the freight train and the
passenger train before they run into each other and we do nothing about it,
then I believe this body is also culpable and negligent.

This idea echoes Aristotle, who held
that ethical decisions are a matter of choice and only ignorance of facts or
lack of freedom to act excuses a person from choosing the ethical action. 22
Senator Feinstein deprived the senators of either excuse. But many Senate
Republicans were unmoved and still tried to stop the bill, believing it imposed
a net economic burden on society. Their effort to thwart its passage with a
filibuster was defeated, and on October 1,2008, just 19 days after the Metro
link accident, the Rail Safety Improvement Act of 2008 became law. 23 The roll
call was 74 to 24. Every Democrat voted for it and all the “nay” votes were
Republicans. These are the main provisions of the 123-page statute.

Mandatory installation by 2015 of positive train
control on rail lines shared by freight and passenger trains, on “main lines”
carrying more than 5 million tons of freight yearly, and on any stretch of
track carrying substances such as ammonia and chlorine that pose toxic inhalation
hazards.

•Rules designed to prevent crew
fatigue, including prohibition of train crews working more than12 hours a day
or 266 hours a month

.•A long list of new mandates for
the Federal Rail-road Administration including certifying conductors,
monitoring locomotive radio traffic, and studying the safety of antique
locomotives used for rides at railroad museums.

•Measures to improve safety at
railroad–highway crossings. Assistance to families of victims of passenger train
accidents.

•A program of annual $50 million
grants to rail-roads for safety improvements.

REGULATORS GO TO WORK

Like many laws passed by Congress,
the Rail Safety Improvement Act is a mixture of specifics and generalities. It
was very precise in dictating work-hour rules for train crews under varying circumstances,
even prohibiting companies from telephoning or paging crew members at home
during mandatory10-hour rest periods. Yet it also set broad new requirements
such as positive train control that left much to the discretion of the Federal
Railroad Ad-ministration. In fact, it gave the agency so much to do it
authorized hiring 200 new employees. Quickly, the agency went to work. Within a
week of the bill’s passage it issued an emergency order prohibiting use of
wireless electronic devices in locomotive cabs and elsewhere on or near
operating trains. 24 It cited seven accidents besides theMetro link collision
where cell phone use distracted engineers. Two led to fatalities. It also
listed examples of unsafe behavior observed by its staff. Some of the stories
were incredible.

An FRA deputy regional administrator
was con-ducting an initial preemployment interview over the telephone with a
locomotive engineer who was applying for an FRA operating practices inspector position.
The deputy regional administrator heard atrain horn in a two long, one short,
and one long pattern and asked the candidate if he was operating locomotive.
The candidate replied that he was, and the deputy regional administrator terminated
the telephone call. The candidate was not selected.

The agency also set to work on a
rule for implementing positive train controls. It began late in 2008 by
convening a working group with representatives from18 organizations including
railroads, unions, suppliers, and the FRA. This group met five times. Between
meetings it broke into task forces. Disagreements between participants were
resolved by FRA decisions. The agency also began a new benefit–cost study.

Within six months it submitted a
proposed rule tothe Office of Information and Regulatory Affairs(OIRA) along
with its benefit–cost study. The 167-pagestudy revealed a stunning excess of
costs over bene-fits. Depending on net present value assumptions, the costs of
positive train controls over 20 years were estimated at between $10 billion and
$14 billion. The safety benefits were only $608 million to $931 million. Under
either assumption the cost of controls was more than 15 times the benefits.

Although
OIRA’s job is to make sure regulations have a net benefit for society, its
hands were tied be-cause of the congressional mandate. It approved the proposed
rule and the FRA published a 79-page No-tice of Proposed Rulemaking in the
Federal ter. 26 This opened a 30-day comment period. Written comments from
anyone could be entered on the FederalRulemaking Portal mailed, faxed, or
hand-delivered to the agency. During this comment period the FRA also held a
one-day hearing at a Washington hotel to give railroads, unions, and state
transportation officials a chance to comment before regulators. When all
comments were in, the agency reconvened its working group to review them and
con-sider changes to the proposed rule. It took six more months. The final rule
was published in the Federal ter on January 15, 2010, and later entered in
the Code of Federal Regulations. 27 It set standards for the design,
functioning, certification, and maintenance of positive train control systems.
It also responded to comments. For example, large railroad companies objected
to a requirement for dual displays in locomotive cabs for both engineers and
brakemen. The agency responded that both were necessary to ensure safety.
General Electric, which sells equipment to the railroads, objected to the agency’s
insistence on approving entire systems and asked it to approve individual parts
or components instead. The agency rejected this suggestion as complicating and
more expensive. Chemical shippers asked to exclude rail lines from controls if
they carried fewer than 100 tank cars of toxic chemicals a year. The agency
refused, saying that was contrary tothe safety mission Congress had given
it.The final rule also revised the 20-year benefit–cost projection, making it
even less favorable. Depending on net present value assumptions, the costs
would be$9.6 billion to $13.3 billion and the benefits $440 million to $674
million, a ratio of more than 20:1 in either case.

WORTH IT?

America now has another expensive
regulatory pro-gram, one that will raise shipping rates, consumer prices, and
rail passenger fares. Railroads are now more heavily regulated in their
operations and employee relations. The Federal Railroad Administration grows
larger and more powerful. On the other hand, rail passengers are safer, and the
railroads may see some efficiency gains. Was the Rail Safety Improvement Act of
2008 justified? Trains are dangerous. Exhibit 1 shows an annual total of
between 700 and 950 railroad fatalities over the past decade, but few of them
were passengers killed in train accidents—only 85 over the 10-yearperiod. Most
fatalities are trespassers who ride trains or enter track corridors. Hundreds
more are motorists hit at crossings. In 2009, for example, there were 713 train
fatalities. Of these only three were passengers killed in accidents. Of the
rest, 446 were people trespassing on tracks, 28 248 were motorists at rail
crossings, and 16 were on-duty railroad employees. A nationwide system of train
controls might have saved the three passengers and some of the 16 railroad
workers killed on duty, but would have done nothing to save the other 697 people.
In absolute numbers, motor vehicles kill far more people than passenger trains,
to be exact 33,960 more in 2009, but they are safer per mile traveled. In
2009the fatality rate per 100 million miles traveled was6.07 for passenger
trains compared with 1.10 for motor vehicles.

In the FRA’s benefit–cost
calculations, the safety benefits of positive train controls were $440 million
to $674 million over 20 years. It assumed that train controls would lead to a
60 percent reduction in rail accident costs including casualties, train delay,
emergency response, and track and equipment damage. 30 A statistical life was
valued at $6 million. The study also noted other potential benefits but did not
monetize them. Possible business benefits for the railroads include the ability
to run more trains, greater reliability, and diesel fuel savings; a possible
benefit for society is reduced pollution from diesel exhaust. However, the
agency concluded that such benefits were uncertain and, even if they appeared,
would not fill the gap between costs and benefits for 20 to 25 years. For
example, trains may have to run more slowly for years as systems are
introduced. 31 The expense to railroads is greater because final cost estimates
for positive train controls were be-tween $9.6 billion and $13.3 billion over
20 years. These include primarily wayside and locomotive components and continuing
maintenance. For example, the railroads must equip roughly 30,000 locomotives
at an estimated $55,000 each. The FRA also predicted that the railroads will
spend 1,729,848hours each year completing new paperwork requirements. 32
Altogether this is a formidable cost burden and an expansion of the regulatory
burden on an industry that often struggles for profitability, all for asafety
rule where, as the FRA concludes, “the cost-to- benefit comparison . . . is not
favorable.” Congress required swift installation of train controls on roughly
69,000 miles of track. Suppliers such as Lockheed Martin and General Electric
have to rush component design. Railroads must make massive, unanticipated
shifts in capital expenditures. In the past they were criticized for rejecting
investments in train controls because capital returns were higher for expenditures
on mergers and track equipment. Now there is no choice. One company, CSX
Transportation, expects to spend $1.2 billion to comply with the new rule.

MEAN WHILE AT METROLINK

Metro link made changes after the
accident. It issued an emergency order against crew use of electronic devices,
added a second engineer in locomotives, in-stalled brighter signal bulbs, and
reduced speeds in some zones. It enlarged its supervisory structure, adding
four new managers to oversee operations and rules compliance, and it asked
Connex to add a new vice president for safety. Efficiency testing of crews was
stepped up. Eventually, it replaced its chief executive and ended its contract
with Connex, giving its operations to Amtrak instead. In 2009 Metro link
installed automatic braking equipment at 43 locations along its routes. This is
an interim measure until it puts positive train controls in place, which it has
agreed to do by 2012, three years before the federal 2015 deadline. It also put
inward-looking video cameras in its locomotive cabs, making it the only
railroad to respond to the NTSB’s recommendation. The Brotherhood of Locomotive
Engineers and Trainmen, which represents Metro link engineers, sued but could
not stop the action. The union said cameras were “punitive in nature” and
breached the right to privacy found in the California constitution. 35 It
recommended equipment to jam cell phones instead, but Metro link continues to
use the cameras. In 2010 Metro link took delivery of 117 new coaches designed
to protect passengers by absorbing energy.

Since the Chatsworth Station
accident there have been no Metro link passenger fatalities from train
accidents.

Questions

1. What were the causes of the Metro link
accident?

2. What could have been done to
prevent the accident? Was management deficient? Were regulators deficient?
Should either have been doing anything differently?

3. Is the cost of positive train
control justified by the likely safety gains for passengers?

4. Did the Federal Railroad
Administration fairly value a statistical life at $6 million?

5. Is money spent to regulate
railroad safety being spent in the most efficient way to reduce risks of death
and injury in society?

6. Were the railroads justified in
opposing legislation to mandate train controls?

7. Is video recording in locomotive
cabs an invasion of privacy? Should unions oppose it?