Stopping the Runaway Train - The Case for Privatizing Amtrak - Cato Institute
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
No. 712 November 13, 2012
Stopping the Runaway Train
The Case for Privatizing Amtrak
by Randal O’Toole
Executive Summary
When Congress created Amtrak in 1970, trak’s, and by 2025 air travel is likely to use less
passenger-rail advocates hoped that it would energy per passenger mile than trains.
become an efficient and attractive mode of Recent efforts to reform Amtrak have proven
travel. More than 40 years of Amtrak operations futile as entrenched interests, ranging from la-
have disappointed them, as Amtrak has become bor unions to local activists, have kept Amtrak
the highest-cost mode of intercity travel and an inefficient carrier heavily dependent on gov-
remains an insignificant player in the nation’s ernment subsidies. No amount of reform will
transportation system. Nationally, average Am- overcome the fundamental problem that, so
trak fares are more than twice as much, per pas- long as Amtrak is politically funded, it will ex-
senger mile, as airfares. Despite these high fares, tend service to politically powerful states even if
per-passenger-mile subsidies to Amtrak are those states provide few riders.
nearly nine times as much as subsidies to air- Budgetarily, Amtrak has become a runaway
lines, and more than 20 times as much as sub- train, consuming huge subsidies and providing
sidies to driving. When fares and subsidies are little or no return. Four decades of subsidies to
combined, Amtrak’s costs per passenger mile passenger trains that are many times greater
are nearly four times as great as airline costs. than subsidies to airlines and highways have
Partly because of these high costs, Amtrak failed to significantly alter American travel hab-
is an insignificant mode of travel. The average its. Simple justice to Amtrak’s competitors as
American flies close to 2,000 miles a year and well as to taxpayers demands an end to those
travels by car around 15,000 miles a year but subsidies.
rides Amtrak only about 20 miles a year. The only real solution for Amtrak is priva-
Nor do the environmental benefits of pas- tization. Private operators would enjoy sub-
senger trains justify Amtrak’s subsidies. Buses stantial cost savings over Amtrak and would be
use far less energy per passenger mile than Am- free to serve those routes that attract the most
trak. Cars in intercity travel are typically as en- passengers rather than the ones that are backed
ergy efficient as Amtrak. While Amtrak is cur- by the most political muscle. Private railroads
rently more energy efficient than flying, airline would also be more likely to develop innova-
energy efficiency is improving faster than Am- tions that will attract new riders.
Randal O’Toole is a senior fellow with the Cato Institute and author of Gridlock: Why We’re Stuck in
Traffic and What to Do about It and American Nightmare: How Government Undermines the
Dream of Homeownership.Before Amtrak Introduction other hand, Amtrak carries only about 1 per-
took over cent as many passenger miles as the airlines
When Congress passed legislation cre- and a fraction of a percent of the passenger
the nation’s ating Amtrak in 1970, passenger-train ad- miles that intercity highways carry, leading
passenger trains, vocates hoped that a national network of critics to argue that it is a waste of money.
passenger trains would be operationally A close look at the data reveal that Am-
rail travel cost profitable and attractive to travelers. Instead, trak has failed for two primary reasons. First,
considerably it has required continuous federal subsidies in most markets passenger trains are simply
less than flying. for operations and repeated injections of fed- not competitive against airline or highway
eral dollars for maintenance. travel. Even in the 100- to 500-mile ranges
Today, average In fact, Amtrak’s appetite for federal funds that rail advocates often say are optimal for
per-passenger- appears to be insatiable. Its budget request passenger trains, buses are far less expensive
mile rail fares are for 2013 was 55 percent more than Congress (and far more energy efficient) than trains.
gave it in 2012. In 2010 Amtrak proposed to Second, government control of Amtrak has
more than twice spend $117 billion upgrading its 457-mile saddled it with numerous inefficiencies, in-
average airfares. Northeast Corridor line, a plan whose cost cluding unsustainably expensive labor con-
ballooned to $151 billion by 2012. For com- tracts and political pressure to maintain
parison, the inflation-adjusted cost of build- service on routes that attract few passengers.
ing the entire Interstate Highway System was Recent years have seen several attempts
only about three times this much, yet it is to reform Amtrak in order to reduce its costs
more than 100 times as long as the Northeast and streamline its operations. But these ef-
Corridor and—unlike Amtrak—it was paid forts have proven futile as entrenched in-
for entirely out of user fees. terests, ranging from labor unions to local
Despite its growing cost, Amtrak’s con- activists demanding that government sub-
tribution to the nation’s transportation is sidize trains they rarely ride, have kept Am-
truly miniscule. The average American travels trak an inefficient carrier heavily dependent
about 15,000 miles a year by auto—well over on huge government subsidies. No amount
3,000 of them on interstate highways—and of reform will overcome the fundamental
1,800 to 2,000 miles a year by air, but only problem that, so long as Amtrak is politically
about 20 miles a year by Amtrak. Even in the funded, members of Congress will pressure it
Northeast, the average resident of the Boston, to provide service to almost every state even
New York, Washington, and intermediate ur- if trains attract few riders in some of those
ban areas travels little more than 40 miles a states.
year in Amtrak’s Northeast Corridor trains. The only real solution for Amtrak is priva-
In short, budgetarily Amtrak has become tization. Private operators would enjoy sub-
a runaway train, eating up huge subsidies and stantial cost savings over Amtrak and would
providing little or no return. Over the past be free to serve those routes that attract the
four decades, subsidies for every passenger most passengers rather than the ones that
mile carried by Amtrak have averaged close to are backed by the most political muscle. Pri-
10 times as much as federal, state, and local vate railroads would also be more likely to
subsidies to airlines and more than 20 times develop innovations that will attract new
as much as subsidies to auto drivers. riders.
This has made Amtrak a continuing puz-
zle for legislators, budget cutters, and policy
analysts. On one hand, federal subsidies to The Highest-Cost Mode
Amtrak are considerably smaller than fed-
eral subsidies to airlines or local subsidies to Before Amtrak took over the nation’s pas-
highways, leading rail passenger advocates senger trains, rail travel cost a little more than
to argue that Amtrak deserves more. On the buses but considerably less than flying. It thus
2Figure 1
Average Airline, Rail, and Bus Fares per Passenger Mile
40
35
30
Cents Per Passenger Mile
25
20
15
10
5
0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Air Rail Bus
Source: 2010 National Transportation Statistics, table 3-16, “Average Passenger Revenue per Passenger-Mile.”
Note: As inflation-adjusted airfares declined by 50 percent, Amtrak fares grew by 70 percent. As a result, Amtrak
is now the highest-cost mode of intercity travel.
offered people of all incomes a genuine alter- the company’s operating losses. During an
native mode of transportation. Today, thanks era when most private transportation costs
to Amtrak management, trains have become significantly declined, Amtrak’s dramatic in-
the highest-cost mode of intercity travel, and crease in fares was stunning.
many of them are patronized mainly by the Not all of Amtrak’s trains collect fares av-
well-to-do who can afford the extra time re- eraging 31 cents per passenger mile. Amtrak
quired for trains rather than flying. divides its trains into three groups: Northeast
In 1970, the year before Amtrak took over Corridor trains that connect Boston, New
the nation’s passenger trains, average rail York, and Washington; short-distance corridor
fares were about one-third less than aver- trains that generally serve cities in one or
age airfares—about 18 cents (in today’s pen- two states, usually with state support; and
nies) versus 27 cents per passenger mile.1 long-distance trains that travel through sev-
Four decades of Amtrak management have eral states.
reversed this ratio and more: by 2011, aver- In 2011 Amtrak collected more than 75
age rail fares were 110 percent greater than cents per passenger mile on the Acela trains
airfares—about 28.5 cents versus 13.8 cents in the Northeast Corridor and 42 cents a Buses are less
per passenger mile (see Figure 1).2 passenger mile on other Northeast Corridor expensive
Part of this change was due to a 50 percent trains. Short-distance corridor trains col-
decline in inflation-adjusted airline fares, lect an average of 22 cents a passenger mile, than Amtrak—
which is beyond Amtrak’s control. But part while long-distance trains collect an average often far less
is due to a 70 percent increase in average rail of 17 cents a passenger mile. These averages
fares, which Amtrak says was necessary due are still greater than average airfares; only
expensive—for all
in part to pressure from Congress to reduce 4 of the 46 trains in Amtrak’s accounting lengths of trips.
3On numerous charged fares less than the airline average of the cost happens at the terminuses, longer
routes, buses 13.8 cents per passenger mile.3 trips cost less per passenger mile. In 2001 the
To be fair to Amtrak, airline trips tend to average airline trip was 1,500 miles, while the
are not only less be longer than rail trips and, since much of average rail trip was 250 miles.4 As shown in
expensive but are
faster and more Table 1
Comparative Fares between Various City Pairs
frequent than
Amtrak. Air Miles Air Amtrak Bus
New York-Seattle 2,400 142 266 139
Chicago-Oakland 1,830 170 205 119
Chicago-Los Angeles 1,740 135 156 119
Chicago-Seattle 1,710 126 159 119
Denver-Oakland 954 90 130 99
New York-Orlando 950 96 127 59
Chicago-Denver 885 74 138 99
Chicago-New Orleans 838 105 117 89
Chicago-New York 731 90 97 55
Oakland-Seattle 670 100 100 70
Atlanta-Washington 547 104 108 49
Oakland-Portland 544 71 80 78
Chicago-Memphis 492 159 99 16
Atlanta-New Orleans 424 79 110 9
Oklahoma City-San Antonio 409 87 59 76
Cincinnati-Washington 409 69 70 11
Boston-Washington 398 65 70 16
Denver-Salt Lake City 390 96 75 69
Chicago-Minneapolis 333 54 102 9
Chicago-St. Louis 258 94 25 13
Chicago-Detroit 234 94 32 6
Houston-San Antonio 192 54 33 8
Portland-Seattle 129 83 32 10
Source: Fares gathered on September 24, 2012, for one-way air and rail trips taken on February 12, 2013, and
one-way bus trips taken on October 23, 2012 (most bus fares were unobtainable for February). Airline fares col-
lected from kayak.com; Amtrak fares from Amtrak.com; bus fares gathered from megabus.com, boltbus.com,
and greyhound.com; air miles from webflyer.com.
Note: Bus companies charge lower fares than Amtrak in almost every market, and are often more frequent and
faster than Amtrak trains. Airline fares are also lower than Amtrak on longer routes.
4Table 1, for trips of more than about 800 air domestic air routes carry nearly 100 times
miles, airlines tend to be significantly less ex- as many passenger miles as Amtrak, while
pensive than Amtrak; for trips of 400 to 800 intercity highways carry some 300 times
air miles, the prices are about the same; and as many passenger miles as Amtrak.8 This
for trips under 400 miles, Amtrak tends to means per-passenger-mile subsidies to Am-
be significantly less expensive than air travel. trak are far greater than subsidies to its com-
While Amtrak may be less expensive than petitors.
air travel for relatively short trips, buses are According to the U.S. Bureau of Trans-
less expensive than Amtrak—often far less portation Statistics, after adjusting for in-
expensive—for all lengths of trips. In 2001 flation to 2011 dollars, subsidies to domes-
(the last year for which comprehensive data tic air travel averaged about $14 billion a
are available), bus fares averaged just under year between 1995 and 2007. Considering
13 cents a passenger mile compared with just that the airlines carried an average of more
over 13 cents for airfares and 25 cents for than 500 billion passenger miles a year dur-
Amtrak. Today, the “new model” of bus ser- ing those years, average subsidies work out
vice pioneered by Megabus costs significant- to about 2.8 cents per passenger mile (see
ly less than that.5 At 217 miles, the average Figure 2).9
bus trip is only a little shorter than the av- Using Bureau of Transportation Statis- Domestic air
erage Amtrak trip, so there is no trip length tics’ numbers, highway subsidies over the routes carry
over which Amtrak has a real competitive ad- same time period averaged about $48 billion nearly 100
vantage.6 a year. Highways carried about 4.1 trillion
On numerous shorter routes, including passenger miles per year, for an average sub- times as many
New York–Buffalo, New York–Toronto, New sidy of 1.1 cents per passenger mile.10 While passenger miles
York–Raleigh, Washington–Richmond, Ra- 95 percent of the airline subsidies came from
leigh–Charlotte, Chicago–Minneapolis, and the federal government, all of the highway
as Amtrak,
Chicago–Indianapolis, buses are not only less subsidies came from state and local govern- while intercity
expensive but are faster and more frequent ments. highways
than Amtrak. On most longer routes, airlines By comparison, federal Amtrak subsidies
are also more frequent and, of course, faster over the same time period averaged 25 cents carry some 300
than Amtrak. Amtrak’s only advantage is per passenger mile.11 State subsidies aver- times as many
that it provides a slightly different quality of aged another 2.8 cents. Per-passenger-mile
service from buses or air, giving passengers subsidies to Amtrak were nearly 9 times sub-
passenger miles
more legroom and greater opportunities to sidies to air travel and nearly 22 times subsi- as Amtrak.
move around the vehicle. But both buses and dies to highway travel.
planes can easily match this; for example, Li- In response to concerns about its large
moLiner, a Boston-to-New York bus, contains subsidies, Amtrak claims that its “‘farebox
just 27 seats on a vehicle that would normally recovery,’ i.e., the portion of operating costs
hold around 55 and offers videos, on-board directly covered by ticket revenue, was 79%
food service, WiFi, and other amenities.7 in fiscal year 2011, compared with 76% in
fiscal year 2010.”12 However, it reaches this
conclusion only with a very liberal defini-
The Highest-Subsidy Mode tion of “farebox revenues” and a very conser-
vative definition of “operating costs.”
Fares are just part of the cost of trans- Actual fares collected by Amtrak in 2011
portation, as most modes of passenger totaled $1.851 billion, which was 79 percent
transport receive some government subsi- of $2.344 billion.13 But $2.344 billion isn’t
dies. Amtrak supporters are quick to point enough to cover Amtrak’s labor, operations,
out that subsidies to air and highway travel and fuel costs, much less materials, facilities,
are greater than subsidies to Amtrak. But advertising, or other operating costs.14 Rev-
5Figure 2
Airline, Highway, and Amtrak Subsidies per Passenger Mile
50
45
40
35
Cents Per Passenger Mile
30
25
20
15
10
5
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Air Amtrak Highway
Source: 2012 National Transportation Statistics, table 3-33, “Transportation Revenues by Mode and Level of
Government,” and table 3-36, “Transportation Expenditures by Mode and Level of Government”; Amtrak annual
reports.
Note: While Amtrak subsidies have varied depending on the generosity of Congress in any given year, when
measured per passenger mile, those subsidies have averaged about 9 times greater than airline subsidies and 22
times greater than highway subsidies.
enues become sufficient to cover 79 percent “under generally accepted accounting prin-
of costs only if they include $109 million in ciples, maintenance is considered an operat-
food service revenues—which arguably are ing expense,” not a capital improvement.15 By
passenger revenues even if they aren’t strictly redefining routine maintenance activities as
farebox revenues—and $191 million in state “capital improvements,” and then pretending
subsidies to short-distance corridor trains— that capital improvements don’t need to be
Per-passenger- which aren’t passenger revenues at all. justified by revenues, Amtrak falsely makes it
mile subsidies On the other hand, Amtrak’s calculation appear that it is becoming more solvent.
of operating costs excludes what it calls “cap- For example, in early 2012, Amtrak issued
to Amtrak were ital improvements.” But most of the money a press release bragging that it was request-
nearly 9 times Amtrak spends on so-called capital improve- ing less “federal operating support” from
subsidies to ments is actually for maintenance costs, in- Congress for 2013 than it received in 2012:
cluding overhauling aging locomotives and $450 million instead of $466 million. But it
air travel and passenger cars, replacing worn-out ties, reno- more than doubled its capital improvement
nearly 22 times vating passenger stations, and renewing elec- request from $657 million to $1.435 billion.
trical hardware needed to power its electric This increase, the agency said, “is neces-
subsidies to locomotives in the Northeast Corridor. As sary to move beyond mere maintenance of
highway travel. the Congressional Research Service observes, existing equipment and infrastructure,” ef-
6fectively admitting that most of its “capital mains an insignificant if not a completely ir- In 2011 the
improvements” in recent years were actually relevant mode of travel. In 2007 domestic air average American
maintenance.16 Even true capital improve- travel peaked at more than 2,000 miles per
ments designed to increase rather than just person. At about the same time, per capita flew 1,800 miles,
maintain service must eventually be repaid, highway travel peaked at about 15,600 miles traveled 15,000
preferably by service revenues. per year, about a third of which was intercity
When counting all costs, including fares travel. Amtrak, meanwhile, carried Ameri-
miles by car, and
and subsidies, Amtrak spends 60 cents per cans an average of just 19 miles per person.17 rode Amtrak just
passenger mile, compared with about 16 Since then, because of the recession, per 21 miles.
cents for airline fares plus subsidies to air capita airline travel has declined by about 10
travel. When amortized capital and mainte- percent, or about 200 miles per person, while
nance costs are included, even trains with the highway travel declined by about 500 miles
lowest fares per passenger mile are a consid- per person. Amtrak travel has increased since
erably more expensive mode of travel than then, but not enough to take up the slack: as
flying when subsidies are included. of 2011, the average American rode Amtrak
21 miles a year, or 2 miles more than in 2007
(see Figure 3). Even this is lower than in the
The Insignificant Mode early 1990s, when Amtrak carried the average
American 24 miles per year.
Amtrak brags that its ridership has Even in the Northeast Corridor, where
reached record levels in recent years. Yet, for Amtrak has the greatest presence, it remains
the vast majority of Americans, Amtrak re- an unimportant form of travel. The 43.5
Figure 3
Per Capita Passenger Miles of Travel by Air and Rail
2,000
1,800
1,600
Passenger Miles Per Capita
1,400
1,200
1,000
800
600
400
200
0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Air Rail
Source: 2012 National Transportation Statistics, table 1-40, “U.S. Passenger Miles.”
Note: The average American travels close to 2,000 miles a year by air but only about 20 a year by Amtrak. Not
shown is per capita intercity highway travel, which averages about 5,000 miles per year.
7million people living in this corridor rode ral electrification, rural Internet access proj-
Amtrak Northeast Corridor trains an aver- ects, the Rural Transit Assistance Program,
age of about 42 miles each in 2011, or about and the Essential Air Service program are all
0.2 percent of the miles of mechanized travel based on this same idea. Ironically, the same
taken by the average American every year. people who advocate these ideas are often
Even adding associated trains between New the first to decry government subsidized ur-
York and Albany, New Haven and Spring- ban “sprawl.”
field, Philadelphia and Harrisburg, and Bos- The cost of actually providing everyone
ton and Portland brings the total to less than with equal access to any of these services is
50 miles per capita.18 prohibitive. Even the Postal Service doesn’t
Amtrak brags that it carries more pas- deliver mail to the door of every rural home.
sengers in the Northeast corridor than the The Census Bureau estimates there are more
airlines, but it admits that it only has about than 3,500 urban clusters of more than
6 percent of the corridor’s total intercity 2,500 people in the United States, and Am-
travel market, with highways carrying 89 trak stops at only about 500 of them.
percent.19 The truth is that if Amtrak disap- Nor do Americans expect that they will
peared tomorrow, autos, buses, and planes have ready access to all possible services no
At average could easily take up the slack without any- matter where they live. People’s decisions to
intercity one noticing any significant increase in con- locate in various places recognize tradeoffs:
occupancy rates gestion on highways or at the airports. locating close to jobs often means sacrific-
Passenger-train advocates claim that, be- ing good schools or quiet neighborhoods;
of 2.4 people per cause the Northeast Corridor is so densely locating in a big city may mean higher-pay-
car, cars are more populated, rail lines that carry the average ing jobs but less affordable housing. Ameri-
resident less than 50 miles per year are some- cans move an average of nearly a dozen
energy efficient how vital to the economic health of the re- times during their lifetimes as their prefer-
than Amtrak. gion. Yet they also make the seemingly con- ences and needs change.21 Taxpayers should
tradictory claim that rail passenger service is not have to subsidize people who choose to
equally critical to the health of rural areas. live in areas that do not support every pos-
Amtrak’s long-distance trains, says the Na- sible transportation service.
tional Association of Railroad Passengers,
“bring economically viable mobility to rural
areas and small towns.”20 Amtrak’s Disappearing
Any proposal to eliminate or reduce Energy Advantage
train service to small towns brings out cries
of protest from people who claim this ser- Passenger-train advocates tout the energy
vice is vital to their community. Yet Am- savings offered by trains over planes and au-
trak’s twice-daily (once in each direction) tomobiles. It is true that Amtrak uses about
passenger trains stopping in towns such as 20 percent less energy per passenger mile
Brookhaven, Mississippi; Las Vegas, New than flying. However, Amtrak’s claims of
Mexico; Libby, Montana; Sandpoint, Idaho; energy savings over autos is more question-
and Winslow, Arizona pick up or drop off an able, while buses have a clear energy-saving
average of fewer than eight passengers per advantage over Amtrak. Moreover, under
train (which means fewer than eight round current trends, even flying will be more en-
trips per day to or from each of these cities). ergy-efficient than trains in a few years.
The notion that rural areas need or de- According to the Department of Energy’s
serve subsidized passenger trains is based Transportation Energy Data Book, Amtrak used
on the idea that all Americans, regardless of an average of 2,271 British thermal units
where they live, should have equal access to (BTUs) per passenger mile in 2010, the latest
all goods and services. The Postal Service, ru- year for which data are available. Commer-
8Figure 4
Air, Rail, and Car Energy Consumption per Passenger Mile
12,000
10,000
BTUs Per Passenger Mile
8,000
6,000
4,000
2,000
0
1970 1975 1980 1985 1990 1995 2000 2005 2010
Air Cars Amtrak
Source: Transportation Energy Data Book, 31st Edition, table 2.13, “Energy Intensities of Highway Passenger
Modes,” and table 2-14, “Energy Intensities of Nonhighway Passenger Modes.”
Note: Amtrak uses as much energy per passenger mile as intercity driving when intercity auto occupancies of 2.4
people per car are used. Meanwhile, Amtrak uses less energy per passenger mile than airlines, but given a likely
continuation of recent trends the airlines will be more energy efficient than Amtrak by 2023.
cial airlines used 2,852 BTUs per passenger of 1.55 people per car to calculate BTUs per
mile. However, the Data Book also notes that passenger mile.25 However, the 1.55 num-
airline energy efficiency has been growing at ber is based on urban travel. According to
3.1 percent per year, while Amtrak’s energy a study commissioned by the California
efficiency has grown by only 1.3 percent per High-Speed Rail Authority, cars in intercity
year (see Figure 4).22 If this trend continues, travel tend to carry more people, an average
airlines will be more energy efficient than of 2.4.26 At 2.4 people per car, the average
Amtrak by 2023. car used 2,226 BTUs per passenger mile in
Aircraft energy efficiency is likely to con- intercity travel in 2010, which makes inter-
tinue to grow faster than rail’s. Boeing says city driving more energy efficient than Am-
that its new 787, for example, uses about 20 trak today. Using an occupancy rate of 2.19
percent less energy per seat mile than the for trips over 75 miles, a Congressional Re- Scheduled
planes it replaces.23 By comparison, General search Service study also found that cars use
Electric estimates that its latest locomotive slightly less energy than Amtrak.27 intercity buses
saves just 3 to 5 percent of the energy re- This doesn’t include light trucks (pick- use 60 percent
quired by its predecessors.24 ups, SUVs, full-sized vans), which consumed less energy per
According to the Data Book, the aver- 7,225 BTUs per vehicle mile or (at 2.4 people
age car used 5,342 BTUs per vehicle mile in per vehicle) 3,010 BTUs per passenger mile passenger mile
2010. The data book uses an occupancy rate in 2010.28 However, the latest federal fuel than Amtrak.
9Even at their non- economy standard requires that the average Who Shot the
wartime peak in of all cars and light trucks sold in 2025 get Passenger Train?
54.5 miles per gallon, which is about 2,300
1920, intercity BTUs per vehicle mile.29 Passenger-rail advocates often blame the
trains carried If auto manufacturers reach this stan- decline of American passenger trains on
dard on a straight-line path from the current federal construction of the Interstate High-
the average average, and if the American auto fleet con- way System.33 But a review of history shows
American fewer tinues to be replaced at the historic average that trains began to decline decades before
than 400 miles of about 5.5 percent per year, then by 2015 that time. Passenger trains reached their ze-
the average automobile, including both cars nith in 1920, when more than 20,000 trains
per year, or less and light trucks, will use less energy per pas- per day traveled over the land connecting
than 3 percent senger mile in intercity travel than Amtrak almost every city and town in the United
as many miles uses today. Between now and 2025, the av- States.34 The railroads carried more than 41
erage energy efficiency of autos will grow at billion intercity passenger miles that year,
as the average 3.2 percent per year, or more than twice as plus another 6 billion passenger miles on
American travels fast as Amtrak’s has grown. commuter trains.35
Scheduled intercity buses already use far Many consider the 1920s to be the gold-
by automobile less energy per passenger mile than Amtrak. en age of passenger trains, but it was a gold-
today. According to a 2008 study commissioned by en age only for those who had the gold to
the American Bus Association, intercity buses pay to ride the trains. While 41 billion pas-
use about 60 percent fewer BTUs per passen- senger miles may sound like a lot, it repre-
ger mile than Amtrak.30 The same conclu- sented less than 400 miles per capita, or less
sion was independently reached by the Con- than 3 percent as many miles as the average
gressional Research Service in 1996 and the Americans travels by automobile today. Rail
National Surface Transportation Revenue fares averaged 2.75 cents per passenger mile
and Policy Commission, a group commis- in 1920, which is well over 30 cents in 2012
sioned by Congress to evaluate federal trans- pennies.36 In terms of worker pay, which was
portation policy, in 2007.31 If Congress wants lower relative to costs in 1920 than it is to-
to save energy, it should stop subsidizing Am- day, such fares would be a dollar or more per
trak, which would encourage an expansion of passenger mile at today’s wages.37 Such high
intercity bus service along Amtrak corridors. fares meant that frequent rail patronage was
The above energy data considers only the confined to the upper and middle classes;
energy costs of operating planes, trains, bus- the vast majority of Americans in 1920 were
es, and autos. When full life-cycle costs are members of rural or urban working classes
considered, including manufacturing, infra- who rarely, if ever, traveled by train.
structure construction, and disposal, rails As shown in Figure 5, competition from
are even less efficient than other modes. A autos in the 1920s led to a slow decline to 24
lifecycle analysis by researchers at the Uni- billion passenger miles in 1929 (less than 200
versity of California found that, because rail miles per capita), and a more rapid decline
lines carry so few passenger miles relative to because of the Depression to as low as 12
highways or airlines, construction, manu- billion passenger miles (less than 100 miles
facture, and disposal consumed many more per capita) in 1932 and 1933. Some railroads
BTUs per passenger mile. Specifically, the responded to these declines by introducing
analysis concluded that, over their complete flashy new trains that carried travelers at
lifecycle, passenger rail lines used about 2.5 much higher speeds than ever before.
times as much energy as they used in just By 1939 trains operated by the Burling-
operations, while highway users consumed ton, Milwaukee Road, Northwestern, Penn-
only about 1.6 times as much energy as in sylvania, Santa Fe, Union Pacific, and several
operations.32 other railroads routinely exceeded 100 mph
10Figure 5
Rail Passenger Miles per Capita before Amtrak
700
600
Intercity Rail Miles Per Capita
500
400
300
200
100
0
1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970
Source: Historical Statistics of the United States: Colonial Times to 1970–Part 2, series Q307_Q308.
Note: Claims that federal construction of the Interstate Highway System killed the passenger train are contra-
dicted by the rapid decline in per capita rail riders during the 1920s and after World War II.
and sometimes reached 120 mph. These rail- in 1954, and 100 miles in 1959. Population
roads along with the New Haven, New York growth failed to compensate for this decline,
Central, Reading, Rock Island, and others and by 1955 total intercity rail passenger In 1959 Trains
scheduled trains with average speeds from miles had fallen well below the lowest levels
terminus to terminus greater than 60 mph of the 1920s. magazine
including the time required for all stops No end to this decline appeared to be in argued that
along the way.38 This was more than twice sight when, in 1959, Trains magazine pub-
the average speed of trains prior to this era, lished an incisive, 38-page report titled, “Who
the passenger
and many of the railroads reported a large in- Shot the Passenger Train?” Written by the train was not
crease in ridership in response to the higher magazine’s pithy editor, David P. Morgan, technologically
speeds. the report was the longest and most impor-
By 1940 ridership was up to nearly 20 bil- tant article in the publication’s 50-plus-year obsolete but
lion passenger miles, or 150 miles per capita. history.39 was “shot in
Gas rationing during World War II led to a Morgan argued that the passenger train
huge increase in intercity rail travel, reach- was not “technologically obsolete,” but that
the back” by
ing 90 billion in 1944, or more than 650 “it was shot in the back.”40 Like Agatha government
miles per capita. Christie’s Murder on the Orient Express, the regulators,
Eager to serve travelers, railroads invested villain was not a single assailant but several.
hundreds of millions of dollars—billions in In particular, Morgan blamed the passenger subsidies to
today’s dollars—in new fleets of locomotives train’s decline on competitors,
and passenger trains after the war. But rid- unfair taxation,
ership crashed immediately following the ●● Unions that insisted on antiquated
war, falling below 300 passenger miles per work rules, such as a 1919 rule re- and obsolete
capita by 1947, 200 miles in 1949, 150 miles quiring railroads to pay crews for a labor rules.
11Trains magazine full day’s pay for every 100 miles they “Nobody is being asked to bail out the
suggested that worked (and to pay overtime for any passenger train in the sense of making it
additional miles), which was still in ef- a Federal ward,” said Morgan. “All that is
passenger trains fect in 1959 even though average train asked is simple justice.” Just eight years
could compete speeds had doubled.41 later, an Arizona attorney named Anthony
●● Government regulators whose work Haswell formed the National Association of
only if they could may have been appropriate when rail- Railroad Passengers specifically to lobby to
run 25 percent roads were monopolies but was unnec- make the passenger train a federal ward.
faster than essary now that competition existed. At least one regulation had a much big-
Morgan cited an example of one state ger impact on passenger trains than freights.
highways, but public utility commission that denied In 1951 an Interstate Commerce Commis-
a 1947 federal a full fare increase requested by a rail- sion (ICC) rule went into effect requiring
regulation made roads because “it has been generally railroads to either install expensive signaling
recognized that railroad passenger ser- equipment or to limit the speeds of passen-
this infeasible by vice, as presently conducted, has not ger trains to 79 mph (see shaded box). A few
limiting trains to and cannot be operated at a profit.”42 railroads, including the Santa Fe, installed
●● Subsidies to competition including the equipment, but most judged that the
79 mph. highway subsidies and the predecessor high cost could not be justified by a small
of what is today called the Essential increment in passenger ridership, so they
Air Service program.43 slowed their trains down, limiting their abil-
●● Unfair taxation of the railroads when ity to compete with auto travel.
publicly owned airports and highways The ICC’s 79-mph speed rule did more
paid no taxes.44 harm than good to passenger trains. Freight
●● Railroad managers who found it eas- cars spend most of their travel time being
ier to deal with freight (which Morgan sorted in yards, so there is virtually no ben-
described as a wholesale business) than efit gained by speeding up freight trains
passenger transport (which Morgan above 79 mph. Railroads could only justify
called a retail business). Morgan didn’t the multimillion-dollar cost of installing the
think that many rail managers were de- ICC-mandated signals by the returns from
liberately trying to kill potentially prof- passenger revenues. Since these returns were
itable passenger trains to make room rapidly declining in the 1950s, few railroads
for even more profitable freight trains that did not already have the signals made
(which some passenger-train advocates the effort.
at the time believed), but he did think A few railroads, such as the Burlington
that most rail managers didn’t really and Illinois Central, were able to compensate
understand the passenger business.45 for the slowing of passenger trains where they
had previously exceeded 79 mph by speeding
Morgan admitted that all but the last of them up in other locations where they had
these culprits hindered rail freight opera- previously traveled at well below 79 mph,
tions as much as they did passenger trains, thus allowing them to maintain the same
yet most railroads were able to earn a profit overall times between major cities. Others,
on their freight business. Still, Morgan ar- such as the Union Pacific and Atlantic Coast
gued that fixing these problems would bring Line, were forced to add time to their sched-
many, if not all, passenger trains closer to ules, making them less competitive with au-
profitability. Certainly the resurgence in rail tos. When manufacturers introduced new
freight after the 1980 deregulation suggested technologies, such as General Motors’ light-
that many rail lines that were once thought weight Aerotrain or the Spanish tilting Talgo
marginal could in fact be profitable once train, railroads that were already up against
some of the above issues were addressed. the 79-mph limit were unable to take advan-
12How the 79 mph Limit Became Law
In 1951 the Interstate Commerce Commission (ICC) implemented a rule requiring
that U.S. trains traveling 80 mph or faster be equipped with expensive signaling equip-
ment instead of relying on traditional forms of railroad signaling. The history of the
rule’s adoption is a typical case of sensationalism and exploitation being used to adopt
questionable public policy.
The ICC first proposed the rule in 1946, when a fast passenger train collided with
the rear end of another passenger train that was halted in Naperville, Illinois, killing 47
people. The two Burlington Railroad trains had been running at about 80 to 85 mph
just three minutes apart when a flagman on the lead train thought he saw something
fall off the train and the conductor ordered the train to stop. The engineer of the fol-
lowing train received a yellow signal indicating the need to slow down, but he didn’t ap-
ply the brakes until the leading train came into view, by which time it was too late. The
second train was still going at least 45 mph when it hit the first.
While the ICC used this wreck to justify the new rule, it isn’t likely that the rule
would have made any difference in this case. If speeds had been limited to 79 mph, the
The Interstate
second train would have hit the first train at 39–44 mph, hardly slow enough to have Commerce
saved many lives. If the railroad had installed the signals the ICC wanted, the only differ- Commission
ence would have been that a yellow light would have lit up in the cab of the locomotive
as well as on the signal outside. The most stringent signaling system would have forced used a train
the train to a stop only if it passed a red signal, and by that time it would have already wreck in
collided with the leading train.
The fireman of the following train did not survive the crash, and despite several in-
Naperville,
vestigations of the crash, the engineer never formally testified about what happened. Illinois, to justify
However, an interior cab signal would not have helped, as it seems likely that the engine the 79-mph speed
crew of the following train saw the yellow and the engineer responded not by setting the
brakes but instead merely backing off on the throttle. When two trains were operating limit, but it’s not
closely together, historian Chuck Spinner explains, “it was not uncommon for the engi- likely that this
neer of the trail train to ‘play the yellow,’ or to push the envelope when it came to speed.” rule would have
Engine crews assumed that a yellow signal meant they were running a bit less than three
minutes behind the train in front, so “all the engineer would have to do is back off on prevented the
the train’s speed just a bit” expecting to increase the distance to three minutes and that wreck.
the next signal would be green.46
The Burlington Railroad responded to the accident by requiring that trains operate
at least 15 minutes apart. This would have given the flagman of the first train time to
walk back far enough to warn the second train of the need to stop. Some of the fatalities
were in a prewar streamlined car that had not been built to the Association of Ameri-
can Railroad’s standards for passenger cars. This car had been crushed between two
stronger and heavier cars, and the ICC directed railroads not to operate cars of different
standards in the same train. These changes made sense. The 79-per-hour rule did not.
tage of these new trains by speeding overall this rule had not been passed, and if the rail-
schedules.47 roads had been deregulated earlier and other
By itself, the ICC 79-mph rule didn’t kill hindrances—such as unfair taxation and
privately operated passenger trains. But if subsidies to competing modes—had been re-
13Figure 6
Rail Subsidies per Passenger Mile
35
30
Cents Per Passenger Mile 25
20
15
10
5
0
Source: “Public Funding Levels of European Passenger Railroads,” Amtrak inspector general, 2008, p. 7.
Note: Amtrak subsidies per passenger mile are about the same as typical subsidies in Europe. Only Britain has
significantly lower subsidies than the average for other European countries.
moved, it is likely that passenger service could reforming federal transportation policy.
have remained profitable on many routes for
A 2008 Amtrak many years.
Inspector General Morgan argued that there were three The European Experience
markets for intercity passenger trains. First,
Report found on routes of 100 to 500 miles, trains could American rail advocates often point to
that virtually compete with other modes if they were “at government support of passenger trains in
least 25 percent faster than the highway.” Europe and lament that the same support
all European The ICC rule made it nearly impossible to isn’t provided in the United States. Yet a close
passenger train achieve this goal profitably. Second, Morgan look at European data reveals that passenger
operations thought that business travelers would still be trains in these countries require large public
attracted to overnight trains between cities subsidies that produce negligible benefits.
require 500 to 1,000 miles apart. However, he noted Contrary to claims that many European
significant public this business was marginal because sleeping rail operations break even or earn a profit,
subsidies, but car passengers cost the railroads more than a 2008 Amtrak inspector general’s report
twice as much as coach passengers, yet paid found that virtually all European passen-
those subsidies less than twice the fares. Finally, Morgan ger train operations “require significant
are often thought that longer routes could sell “only public subsidies.” These subsidies are often
on a cruise basis whereby ‘getting there is disguised in rail budgets. For example, Eu-
disguised in half the fun.’”48 Whether these markets are ropean rail operations “are typically orga-
national budgets. still viable today can only be determined by nized into two separate business entities,”
14one to manage the infrastructure and one of communism—traveled 6,459 miles by car, Spain has
to operate the trains. The train operators compared with 14,510 for the average Ameri- invested more in
may report a profit, but only because they can. European subsidies to rail made up for
can ignore the heavy losses experienced by only a small fraction of this gap: while in 2006 high-speed trains
the infrastructure managers. Even the prof- the average American traveled only 18 miles than any other
its reported by the train operators are often by intercity rail and 35 miles by commuter
exaggerated by including government subsi- rail, the average Western European travelled
European nation,
dies among their revenues.49 533 miles by intercity and commuter rail. The yet since it began
The inspector general report estimated the extra miles of rail travel made up for less than building high-
government contributions to passenger op- 6 percent of the auto travel gap.51
erations for each of nine European nations, Americans actually made up for most of speed rail, rail’s
plus the United States, from 1995 through the rail gap with greater bus travel. While the share of Spanish
2003.50 Comparing these subsidies with pas- average EU-15 resident rode buses 665 miles passenger travel
senger miles of rail ridership during those in 2006, the average American rode them
years reveals that the subsidies in most of 997 miles, including airport buses, tour bus- declined from
these countries are between 20 and 30 cents es, and school buses, as well as urban tran- 6.9 percent to
per passenger mile, with the United States sit and intercity buses. Europeans also flew
being right in the middle (see Figure 6). The less than Americans. While separate data
5.4 percent.
only major exception is the United Kingdom, are not available for the EU-15, the average
where subsidies are well under 15 cents per residents of the EU-27 (which includes the
passenger mile, partly because it privatized original EU-15 as well as Cyprus, Malta, and
its rail operations in the mid-1990s. This sug- 10 former communist countries) flew about
gests that there are no economies of scale that 700 miles in 2006, compared with 1,970
would allow Amtrak to reduce its per passen- miles by Americans.52
ger subsidies by expanding to a much greater America’s greater mobility is not due to
size. The total subsidies in the nine countries the nation’s large land area or low popula-
during the years studied averaged $38 billion tion densities. The second-most mobile na-
per year, while annual subsidies to Amtrak in tion in the world, after the United States, is
those years averaged only about $1 billion. Iceland, which has only 1 percent of the land
The average resident in these nine countries area of the United States.53 A comparison of
rode trains about 34 times as many miles as Western European nations reveals almost
the average American, suggesting that there no correlation between population densities
are diminishing returns to passenger rail sub- and miles of per capita travel.54
sidies: a doubling of subsidies will produce Nor is European spending on trains dra-
less than a doubling of ridership. matically increasing rail’s share of European
Together, the nine countries studied by travel. According to the European Union, in
the inspector general have a slightly greater 1980 76.4 percent of European travel was by
population than that of the United States, auto, 2.5 percent was by air, while 8.2 percent
but less than one-fourth the land area, re- was by intercity train.55 By 2006 auto’s share
sulting in an average population density that was still 76.4 percent, while air’s share had
is 4.5 times greater than the United States. grown to 8.3 percent and rail’s share had de-
But this higher density has minimal effect clined to 7.9 percent. The big loser was buses,
on European travel choices, which are more whose share declined from 11.9 percent to
influenced by the high taxes European gov- 6.3 percent.56
ernments impose on motor fuel, which effec- Spain has invested more in high-speed rail
tively suppress European mobility. than any other European nation and now has
In 2006 the average resident of the EU- more miles of high-speed rail than any coun-
15—the 15 Western European nations that try in the world except China. In 1990, before
made up the European Union before the fall the nation’s first high-speed line opened, rail’s
15share of surface passenger travel in Spain was transport by nearly 50 percent, while rail’s
6.9 percent. By 2010, after a total of four ma- share declined in the rest of the EU-15.62 The
jor high-speed rail lines had opened, rail’s private operators accomplished this growth
share had declined to 5.4 percent.57 with less than 60 percent of the average sub-
The most interesting passenger-rail ac- sidy per passenger mile of the eight other
tion in Europe has been the British attempt national rail systems studied by the Amtrak
to reform its rail system. This reform is of- inspector general in 2008.
ten described as privatization, but it really Three major complaints about the Brit-
ended up being a form of contracting. Ini- ish railway program had to do with safety,
tially, Britain sold its rail infrastructure to a reliability, and cost to taxpayers. However,
private company, called Railtrack, that was since Network Rail replaced Railtracks, both
supposed to maintain the rail lines and allow safety and on-time performance have im-
any operator to use them provided they paid proved. In 2010, 92 percent of trains were
an access fee. The government determined on time, and surveys show that 83 percent
where passenger trains should run and invit- of customers are satisfied with rail services.
ed private operators to bid on those routes. The private operating companies have pur-
The Railtrack portion of the program chased new equipment, reducing the aver-
Britain’s system proved to be a failure. In an effort to save age age of rolling stock by 23 percent.63
of contracting money, it deferred maintenance, resulting in The biggest complaint about the British
passenger a poor on-time record as well as three seri- program was that subsidies, which were sup-
ous accidents that killed 42 people between posed to decline, actually increased. Adjust-
trains to private 1997 and 2000. As a result, the government ing for inflation, subsidies averaged about 2
operators led bought out Railtrack and now operates the billion pounds per year before privatization.
rail infrastructure through a quasi-govern- These subsidies actually did decline slightly
to a 60 percent mental organization called Network Rail.58 under Railtracks, but since Network Rail
increase in rail Private contractors were allowed to bid took over the infrastructure, they have grown
travel between on various routes, and if the bids were nega- to more than 4 billion pounds per year. The
tive—that is, if the contractors would require biggest increase in costs was infrastructure
1996 and 2010, subsidies to operate the routes—then the renewals, an increase partly required by
far more than government would provide those subsidies. Railtrack’s deferral of maintenance.64 Since
The private companies own their own loco- overall costs per passenger mile remained
in any other motives and railcars, but they operate under constant, much of the increased subsidy can
European franchise to the government. be attributed to the growth in rail ridership.
country. Some reviewers consider the British sys- Some argue that the government re-
tem a failure.59 Yet in one important way— purchase of rail infrastructure proves that
ridership—it has been a huge success. From privatization doesn’t work. But a more valid
1996, when private takeover was largely com- interpretation is that the British system con-
pleted, to 2010, British rail passenger miles tinued to rely too much on central planning
grew by 60 percent, far more than in any oth- and regulation. Instead of letting the market
er European country, and more than twice as determine where passenger service would
much as in the EU-15 as a whole.60 be by rail, bus, or another mode, the Brit-
Rail’s share of British passenger travel ish government politically determined rail
grew by 70 percent from 4.4 percent in 1995 routes and agreed to subsidize the suppos-
to 7.5 percent in 2010. Again, this was more edly privatized rail operators. The division of
than that of any other European country; the system between an infrastructure com-
rail’s share of travel in the EU-15, minus pany and operating companies failed to cre-
Britain, grew by only 10 percent in the same ate the proper incentives, which contributed
time period.61 During the same years, private to safety and reliability problems. Despite
operations also increase rail’s share of freight these failings, no other country, no matter
16how much it invested in high-speed rail or To reform Amtrak, some Republicans
other rail improvements, has come close to a have proposed to have it contract out its op-
70 percent increase in rail’s share of passen- erations to private companies such as Veolia
ger travel. and Stagecoach, both of which operate ur-
While not attempting widespread privati- ban rail transit lines in the United States and
zation, other European countries are saving elsewhere. Contracting out might help solve
money by contracting out selected routes. the labor problem, except that the same pro-
Contracting has saved 20 to 50 percent of posal calls for a requirement that contractors
costs on routes in the Netherlands, 20 to 30 pay employees as much as Amtrak is paying
percent in Sweden, and 20 percent in Ger- them. Contracting out could also encourage
many.65 operators to be efficient, especially so long
as there was competition for such contracts.
However, contracting out would not end the
Privatization or inefficiencies that result from political deter-
Contracting? mination of routes and services.
Many American transit agencies contract
Amtrak is supposed to be a corporation, out their operations at considerable savings
not a government agency, and Amtrak presi- to taxpayers. The Denver Regional Tran-
dent Joseph Boardman says that it is his goal sit District (RTD) is a good example, as the
to run Amtrak like a business. But there are Colorado legislature required the agency to
three obstacles to running a government- contract out half of its bus routes. The com-
funded organization like a business. panies that operate the routes maintain and
First, an organization that can draw on fuel their buses, pay taxes on their fuel and
tax dollars to cover losses has little incentive property that RTD is exempt from, and pay
to operate efficiently. Everyone in the organi- workers roughly the same amount as RTD.
zation, from top managers on down, knows One of the companies is even unionized.
that they can fall back on tax dollars to make Despite higher fuel costs (when taxes are in-
up for any inefficiencies in their operations. cluded) and comparable labor costs, in 2010
Second, an organization that is beholden the private contractors cost 49 percent as
to government to fund its operations must much per bus-revenue mile and 52 percent as
do things that are pleasing to government. much per bus-revenue hour than the buses
Since Amtrak gets a third to half its funds operated by RTD.66
from tax dollars, it must please appropria- Judging from experiences in Europe,
tors as much as its customers. Among other contracting out Amtrak’s operations might
things, this means running trains that car- reduce per-passenger-mile subsidies by 20
ry few passengers simply because they go to 50 percent. This would allow Congress
through the states or districts of powerful to reduce Amtrak’s appropriations without Complete
members of Congress. threatening Amtrak’s routes or allow Am- privatization
Third, Amtrak faces the same labor prob- trak to operate more trains on the same an-
lem as other public agencies in that nego- nual appropriation.
of Amtrak is
tiations with labor unions are asymmetric. Contracting out remains an imperfect the only way to
When private companies negotiate with solution because Amtrak’s route map would avoid the waste
unions, the negotiators clearly represent still be politically drawn rather than focused
each side of any dispute. But when public on the most suitable markets for passenger that results
agencies negotiate, the agency negotiators trains. Amtrak would still need to persuade from political
have a conflict of interest as they represent Congress to subsidize most routes, and to do determination
both taxpayers, who have to pay for what- so it would need to extend service into most
ever the two sides agree to, and union mem- states even if some provided few customers. of routes and
bers, who are also voters. In the long run, both Amtrak and the pri- services.
17To be fair, vate operators would lack market incentives serve remote communities. It should also
Amtrak to provide the most cost-effective transpor- encourage state and local governments to
tation service. The proposal to require con- find ways to fund highways entirely out of
privatization tractors to pay workers as much as Amtrak user fees rather than general tax dollars.
should take place pays also represents an inefficient and inap- Though privatization might result in
propriate meddling with the market. an end to passenger trains on some routes,
in the context of Complete privatization of Amtrak is the many routes could survive. Table 2, which
reforming other only way to avoid these problems. Such shows data for the 44 routes in Amtrak’s sys-
transportation privatization could be part of a compre- tem, provides some clues as to what routes
hensive transportation reform package that those would be. The table shows the “op-
programs, such eliminates subsidies to rail competitors, erating profit” for each route in 2011; this
as ending federal reduces whatever regulatory constraints re- includes state subsidies on the revenue side
subsidies to main on private operators, and provides for but does not include depreciation, interest,
fair taxation of all modes of transportation. capital improvements, or maintenance on
airlines and Privatization would mean selling Am- the cost side. “Cost per passenger mile” is
encouraging trak’s assets: 623 route miles between Boston equal to the revenue per passenger mile (not
and Washington, New Haven and Spring- shown) minus the profit per passenger mile.
state and local field, Philadelphia and Harrisburg, and in The table shows only four routes that
governments to Michigan; 105 rail stations; maintenance made an operating profit in 2011: the Acela
fund highways facilities; and roughly 2,300 locomotives, and regional trains in the Northeast Corridor;
passenger cars, and other rolling stock. Am- the New York-Montreal Adirondack; and the
out of user fees trak estimates that these and other assets are Washington-Lynchburg trains (which are an
rather than worth about $11.1 billion, while its liabilities extension of the Northeast Corridor). These
are about $5.3 billion.67 trains earned an operating profit either by at-
general tax These numbers, however, represent the tracting patrons willing to pay high fares (as
dollars. costs to Amtrak of purchasing right of way, in the Northeast Corridor trains) or keeping
infrastructure, and rolling stock, minus de- costs low (as in the other two trains). Keep
preciation. Just because Amtrak paid these in mind that when capital improvements
amounts to provide trains that lose money, and maintenance are included, none of these
however, doesn’t mean they are worth that trains made a profit, and those costs are prob-
much on the open market. Privatization will ably higher in the Northeast Corridor (which
probably produce net revenues that are less is owned and must be maintained by Amtrak)
than the $5.8 billion Amtrak claims is its net than on most other routes, where Amtrak
asset value. pays private railroads an access fee (which is
Privatization should be done with a mini- included in operating costs) and need only
mum of restrictions or regulation. Whoever maintain its rolling stock and stations.
buys Amtrak’s routes should not be required The table reveals a very wide range of op-
to open those routes to any train operator. erating profits and costs per passenger mile
Other than meeting basic safety require- as well as occupancy rates (that is, the average
ments, operators should not be required to percentage of seats filled by revenue custom-
meet any particular frequency or level of ser- ers). Operating profits range from 25 cents
vice. Passenger trains, like buses and planes, per passenger mile for the Acela to minus 47
should respond to the market, not to politi- cents per passenger mile for the Los Angeles-
cians. New Orleans Sunset Limited. (The even greater
To be fair, such privatization should take loss of 80 cents a passenger mile on the Chi-
place in the context of reforming other parts cago-Indianapolis train is mitigated by the
of the transportation system as well. Con- fact that Amtrak uses this train primarily as
gress should end the Essential Air Service a “hospital train” to transfer cars in need of
subsidies and requirements that airlines service from its Chicago hub to its shops in
18You can also read
