This is the accessible text file for GAO report number GAO-02-398 
entitled 'Intercity Passenger Rail: Amtrak Needs to Improve Its 
Decisionmaking Process for Its Route and Service Proposals' which was 
released on April 22, 2002. 

This text file was formatted by the U.S. General Accounting Office 
(GAO) to be accessible to users with visual impairments, as part of a 
longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the 
printed version. The portable document format (PDF) file is an exact 
electronic replica of the printed version. We welcome your feedback. 
Please E-mail your comments regarding the contents or accessibility 
features of this document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

United States General Accounting Office: 
GAO: 

Report to the Honorable Ron Wyden: 
U.S. Senate: 

April 2002: 

Intercity Passenger Rail: 

Amtrak Needs to Improve Its Decisionmaking Process for Its Route and 
Service Proposals: 

GAO-02-398: 

Contents: 

Letter: 

Results in Brief: 

Background: 

Status of the Growth Strategy: 

Amtrak Overestimated Expected Mail and Express Revenue: 

Amtrak Encountered Substantial Difficulties in Expanding Service Over 
Freight Railroad Tracks: 

Conclusions: 

Recommendation for Executive Action: 

Agency Comments and Our Evaluation: 

Scope and Methodology: 

Appendix I: Financial Performance of Amtrak’s Routes, Fiscal Year 2001: 

Appendix II: Amtrak Route Actions, January 1995 Through December 2001: 

Appendix III: Planned Route and Service Actions Included in the 
Network Growth Strategy: 

Appendix IV: Amtrak’s Process for Evaluating Route and Service 
Proposals: 
Amtrak’s Consideration of Operating Revenue and Direct Costs: 
Consideration of Capital Costs and Other Financial Issues: 

Appendix V: Market-Based Network Analysis Models Used to Estimate 
Ridership, Revenues, and Costs: 
Models Used to Estimate Ridership and Revenue: 

Models Used to Estimate Costs: 

Appendix VI: Comments from the National Railroad Passenger Corporation: 
GAO’s Evaluation: 

Tables: 

Table 1: Status of Network Growth Strategy Route and Service Actions, 
as of December 31, 2001: 

Table 2: Operating Profit (Loss), Operating Ratio, and Profit (Loss) 
per Passenger of Each Amtrak Route, Fiscal Year 2001, Ranked by Profit 
(Loss): 

Table 3: Planned Network Growth Strategy Route and Service Actions: 

Figure: 

Figure 1: Amtrak’s Route System, as of December 2001: 

[End of section] 

United States General Accounting Office: 
Washington, DC 20548: 

April 12, 2002: 

The Honorable Ron Wyden: 
United States Senate: 

Dear Senator Wyden: 

The National Railroad Passenger Corporation (Amtrak) is the nation’s 
intercity passenger rail operator. In recent years, facing the 
continuing deterioration of its financial condition and its attempt to 
eliminate its need for federal operating assistance by December 2002, 
Amtrak undertook a number of actions intended to improve its financial 
condition.[Footnote 1] Among these actions were changes in its routes 
and services.[Footnote 2] Starting in 1995, Amtrak began reducing 
service and eliminating routes to reduce costs, although this strategy 
was ultimately unsuccessful. In December 1999, Amtrak’s board of 
directors shifted its route and service strategy toward planning to 
implement new routes and expand services (called the Network Growth 
Strategy) on the freight railroad tracks over which Amtrak operates. 
It estimated that this expansion, involving 15 routes, would yield 
about $66 million in net financial benefits through fiscal year 2002, 
primarily from expanding its mail and express business, and help 
Amtrak move toward operational self-sufficiency.[Footnote 3] 

In response to your interest, we are reporting on (1) the status of 
the Network Growth Strategy, (2) Amtrak’s estimates of expected mail 
and express revenue, and (3) Amtrak’s success in obtaining freight 
railroads’ agreement to allow it to expand service over their tracks. 
To carry out our work, we reviewed documents describing the Network 
Growth Strategy, among other things. We also reviewed Amtrak’s 
financial information used to generate route and service actions. We 
discussed the proposed Network Growth Strategy actions with officials 
from Amtrak, the Department of Transportation, and several freight 
railroads over which Amtrak operates. In February 2002, after we 
completed our work, Amtrak estimated that it would need more than 
twice the amount of federal appropriations ($1.2 billion for fiscal 
year 2003) than it had received on average over the past 3 years to 
retain its route network. We did not assess the reasonableness of this 
estimate. 

Results in Brief: 

Amtrak has not been successful in implementing its Network Growth 
Strategy. About 2 years after announcing it, Amtrak has only 
implemented three routes, one of which was later canceled. Nine routes 
were canceled without being implemented. Amtrak still plans to 
implement the remaining three routes, although later than planned. 
Typically, more than one reason—-such as overestimation of express 
revenue and inability to reach agreement with freight railroads over 
various aspects of implementing the proposals—-contributed to the 
cancellation of any single proposed route or service action. 

Increased mail and express revenues were the cornerstone of the 
Network Growth Strategy. However, Amtrak overestimated the mail and 
express revenue it expected to obtain. Amtrak’s assessment of the $66 
million in net financial benefits it expected to obtain was 
speculative because it was not based on a rigorous analysis of the 
express business that it might expect to receive. According to Amtrak, 
this overestimation occurred because (1) it had no empirical basis for 
its revenue estimates and (2) express shippers were reluctant to enter 
into contracts for service that did not yet exist. Six of the planned 
route actions were canceled in part because Amtrak overestimated the 
mail and express revenues associated with these routes. According to 
Amtrak, it has since revised its revenue estimating process to focus 
more on determining existing customers’ shipping needs, assessing 
these needs in light of current economic trends, and evaluating 
Amtrak’s ability to meet these needs given existing train capacity. 

For the most part, Amtrak was not able to reach agreement with freight 
railroads over whose tracks it would operate. Amtrak told us that it 
expected that freight railroads would provide capital funds needed to 
implement some routes, which the freight railroads were not inclined 
to do. Freight railroads were also concerned about (1) Amtrak’s plans 
to operate additional trains in areas that were already congested, (2) 
Amtrak’s plans to carry express merchandise that might compete with 
their own business, and (3) compensation that Amtrak would pay them 
for use of their tracks. Even though operating over freight railroad 
tracks was a crucial component to the Network Growth Strategy’s 
success, Amtrak did not obtain a full understanding of freight 
railroad concerns before deciding to implement the strategy and before 
announcing it to Congress. Amtrak canceled six routes, in part, 
because it could not reach agreement with freight railroads over these 
issues (including four routes that were canceled in part because of 
the overestimation of mail and express revenues). We are making a 
recommendation to improve Amtrak’s process for making future route and 
service decisions. 

In commenting on a draft of this report, Amtrak agreed with our 
presentation of the reasons that it overestimated mail and express 
revenue. Amtrak disagreed with our assessment that it was largely 
unsuccessful in obtaining freight railroads’ agreement with its 
strategy because it did not fully understand significant 
implementation concerns that freight railroads might have had before 
it decided to implement the Network Growth Strategy. We have clarified 
our report and revised our recommendation to make it more useful to 
Amtrak. The Department of Transportation generally agreed with the 
conclusions of our draft report. 

Background: 

Amtrak was established by the Rail Passenger Service Act of 1970, 
after the nation’s railroads found passenger service unprofitable. It 
operates a 22,000-mile passenger rail system over 43 routes. (See 
figure 1.) Amtrak owns 650 miles of track, primarily in the Northeast 
Corridor, which runs between Boston, Massachusetts, and Washington, 
D.C. It operates the remainder of its routes over tracks owned by 
freight railroads and pays these railroads for this access. From 
fiscal year 1971 through 2002, the federal government provided Amtrak 
with about $25 billion in federal operating and capital assistance, or 
an average of $807 million annually (in nominal dollars). 

Figure 1: Amtrak’s Route System, as of December 2001: 

[Refer to PDF for image: illustrated U.S.map] 

Map depicts: 
Amtrak Train Routes; 
Planned New Amtrak Train Routes. 

Source: Amtrak. 

[End of figure] 

Amtrak’s financial condition has been deteriorating and, in 2001, it 
experienced its largest net loss (revenues less expenses) ever of 
about $1 billion. In fiscal year 2001, only one of Amtrak’s 43 routes 
made enough revenue to cover its operating costs—the Metroliner/Acela 
Express service on the Northeast Corridor ($51 million).[Footnote 4] 
The other 42 routes had operating losses ranging from about $600,000 
on the Vermonter (service between Washington, D.C., and Vermont) to 
$71.5 million on the Northeast Direct (generally service between 
Virginia and Massachusetts). (See appendix I for the financial 
performance of all Amtrak routes.) 

Amtrak has changed its general approach to route and service actions 
over time, from attempting to improve its financial performance by 
cutting service to attempting to achieve the same result by increasing 
service. For example, in 1995, Amtrak eliminated 9 routes, truncated 3 
routes, and changed the frequency of service on 17 routes.[Footnote 5] 
These actions were intended to cut costs by about $200 million while 
retaining 90 percent of revenues and 85 percent of passengers. Amtrak 
said the presumption was that passengers would use other trains to 
meet their travel needs, allowing it to retain most of its ridership 
and revenue. Although initially the route cutting actions had some 
financial success, subsequent financial results were below 
expectations because (1) management did not cut costs as planned, (2) 
less-than-daily service caused less efficient equipment usage and 
other unforeseen problems, and (3) passengers were no longer adjusting 
their travel plans to fit Amtrak’s new less-than-daily schedules. 
[Footnote 6] 

In 1998, Amtrak switched its strategy to increase revenues by 
expanding service. It used a market-based approach to identify the 
market for intercity passenger rail service. To do so, it used market 
research and computer-based models to determine the potential 
ridership, revenue, and costs of proposed route and service actions. 
According to Amtrak, this approach constituted a significant 
improvement in its route evaluation process because it represented the 
first comprehensive analysis of Amtrak’s route system, and the first 
attempt to apply rigorous financial analyses and modeling techniques 
to the design of Amtrak’s national network. 

The Network Growth Strategy was the first product of the market-based 
network analysis project that Amtrak initiated in October 1998 to 
address route evaluation deficiencies. The intent of the market-based 
network analysis was to (1) develop the financial tools that Amtrak 
needed to perform reliable and objective analyses of route and service 
changes; (2) help Amtrak achieve operational self-sufficiency by 
December 2002 by identifying route and service changes that, if 
implemented expeditiously, would produce positive financial impacts 
before the statutory deadline; and (3) express Amtrak’s vision of how 
its national network could be enhanced and improved. 

In December 1999, Amtrak’s board of directors adopted the Network 
Growth Strategy as part of Amtrak’s strategic business plan. The 
strategy consisted of 15 planned route and service actions, the 
majority involving the expansion of service. (See app. III.) Amtrak 
predicated the growth strategy on the acquisition of significant new 
revenue from hauling mail and express cargo and estimated that it 
would result in $65.6 million in net revenue through fiscal year 
2002.[Footnote 7] In February 2000, Amtrak announced to Congress that 
it was going to implement the 15 routes in the Network Growth Strategy. 

Status of the Growth Strategy 

Amtrak has been unsuccessful in implementing its Network Growth 
Strategy. About 2 years after announcing the Network Growth Strategy, 
Amtrak has canceled 9 of the 15 planned route actions without 
implementing them.[Footnote 8] Amtrak implemented three route actions, 
although it canceled one of these in September 2001. Finally, Amtrak 
plans to proceed with 3 other route actions, although their 
implementation will be at least 1 or 2 years later than originally 
planned. (See table 1.) According to Amtrak, the capital funds for one 
of the projects in planning (Silver Service restructuring in Florida) 
were frozen on February 1, 2002, in a company-wide effort to reduce 
use of cash. (In all, Amtrak canceled nine routes without implementing 
them. Some routes were canceled for more than one reason.) 

Table 1: Status of Network Growth Strategy Route and Service Actions, 
as of December 31, 2001 

Name: Hiawatha (extension); 
Planned start of service: Spring 2000; 
Status: Canceled. 

Name: Lake Cities; 
Planned start of service: Spring 2000; 
Status: Canceled. 

Name: Chicago-Janesville, Wisconsin (Lake Country Limited); 
Planned start of service: Spring 2000; 
Status: Implemented in 2000, canceled in 2001. 

Name: Skyline (Manhattan Limited); 
Planned start of service: Summer 2000; 
Status: Canceled. 

Name: Silver Meteor extension to Boston; 
Planned start of service: Summer 2000; 
Status: In planning. 

Name: Crescent Star; 
Planned start of service: Summer 2000; 
Status: In planning. 

Name: Aztec Eagle; 
Planned start of service: Summer 2000; 
Status: Canceled. 

Name: Texas Eagle (restore daily service); 
Planned start of service: Summer 2000; 
Status: Implemented in 2000. 

Name: Twilight Limited; 
Planned start of service: Fall 2000; 
Status: Canceled. 

Name: Luxury Transcontinental; 
Planned start of service: Fall 2000; 
Status: Canceled. 

Name: International (reroute); 
Planned start of service: Fall 2000; 
Status: Canceled. 

Name: Chicago-Des Moines (Hawkeye); 
Planned start of service: Fall 2000; 
Status: Canceled. 

Name: Silver Service restructuring in Florida (including service on 
Florida East Coast Railway); 
Planned start of service: Fiscal year 2001; 
Status: In planning. 

Name: Sunset Limited (reroute in Texas); 
Planned start of service: Fiscal year 2002; 
Status: Canceled. 

Name: Kentucky Cardinal extension; 
Planned start of service: Unspecified; 
Status: Implemented in 2001. 

Source: GAO’s analysis of Amtrak’s data. 

[End of table] 

Amtrak Overestimated Expected Mail and Express Revenue: 

Amtrak told us that it canceled six of the Network Growth Strategy 
routes before they were implemented, in part, because it overestimated 
expected increases to mail and express revenue under the Network 
Growth Strategy. Amtrak estimated that this expected increase would 
improve Amtrak’s bottom line by $65.6 million through fiscal year 
2002. Specifically, it estimated that mail and express revenues would 
exceed costs by $68.2 million, offsetting a loss of $2.6 million from 
expanded passenger operations. Most of the revenue increase was 
expected to come from new express business. This expanded mail and 
express traffic did not materialize and Amtrak’s revised plans have 
reduced expected Network Growth Strategy-associated mail and express 
revenue by about half—-from $271 million to $139 million (a $132 
million reduction). 

Amtrak said that there were several reasons why this overestimation 
occurred. The current president of Amtrak’s mail and express unit told 
us that Amtrak expected to substantially expand its route system to 
generate this revenue and to begin running longer trains mostly filled 
with express traffic. However, he said that at the time Amtrak made 
its mail and express revenue estimates, it gave little thought to 
whether such an expansion was feasible-—that is, whether Amtrak could 
likely capture this business or whether freight railroads that own the 
tracks would agree to Amtrak’s expansion plans. According to Amtrak, 
it did not have a rigorous approach to estimating expected express 
business. Amtrak officials told us that, until recently, Amtrak 
estimated express revenue largely on the basis of an analysis of a 
database of commodities being shipped nationally. Amtrak estimated the 
portion of this business that it thought it could obtain. An Amtrak 
official said that it now focuses more on determining existing 
customers’ shipping needs, assessing these needs in light of current 
economic trends, and evaluating Amtrak’s ability to meet these needs 
given existing train capacity. Finally, Amtrak officials told us that 
express shippers were reluctant to enter into contracts for service 
that did not yet exist. 

Amtrak officials also told us that the company did not know route-by-
route costs for its mail and express program when it announced its 
Network Growth Strategy. This is because Amtrak has never separately 
identified these costs. Rather, it has integrated these costs into the 
overall financial results of its intercity strategic business unit. 
Knowing these costs was important because Amtrak expected that the 
expansion of mail and express service would produce the revenue needed 
to make its route expansion profitable. Not until 2000 did Amtrak 
begin efforts to separately identify mail and express costs and 
develop separate mail and express financial information.[Footnote 9] 
According to Amtrak, in October 2001, it began producing separate 
profit and loss statements for its mail and express business. However, 
an Amtrak official said the corporation still has a long way to go in 
producing reliable mail and express financial information and in 
understanding the true cost of this business. 

Amtrak Encountered Substantial Difficulties in Expanding Service Over 
Freight Railroad Tracks: 

Amtrak could not implement its Network Growth Strategy unless it 
reached agreement with freight railroads over funding for capital 
improvements[Footnote 10] (such as upgrading tracks and signals to 
improve safety and support higher-speed passenger operation) and 
access to freight railroads’ track. Quick agreement was necessary 
because Amtrak wanted to implement the new routes and services to help 
it reach operational self-sufficiency by December 2002. Amtrak 
encountered substantial difficulties in gaining freight railroad 
agreement to allow Amtrak to expand service over freight railroad 
tracks. This difficulty in reaching agreement contributed, in part, to 
Amtrak canceling six of its planned routes. 

Amtrak planned to operate the 15 Network Growth Strategy routes over 
freight railroad tracks, including the transportation of mail and 
express as authorized by law. However, Amtrak was largely unable to 
gain freight railroads’ agreement. Such agreement was critical to the 
implementation of Amtrak’s strategy. Freight railroads are required by 
law to allow Amtrak to operate over their tracks and to give Amtrak’s 
trains priority over their own. In addition, freight railroads are 
required to charge Amtrak the incremental cost-—rather than the full 
cost—-associated with the use of their tracks. These amounts are 
negotiated by Amtrak and the freight railroads. Federal law also gives 
Amtrak the authority to carry mail and offer express service. These 
mandates result in an ongoing tension between Amtrak and freight 
railroads for several reasons. One reason is that accommodating 
passenger trains affects freight railroads’ ability to serve their 
customers and earn a profit. Second, accidents involving passenger 
trains may create liability issues for freight railroads. Third, 
freight railroads believe that they are not fully compensated for 
providing this service. Finally, Amtrak’s express business may compete 
with freight railroads’ business and Amtrak may be able to offer lower 
rates than freight railroads, everything else being equal, because 
Amtrak only has to pay freight railroads the incremental, rather than 
the full cost, of operating on freight railroad tracks. 

According to Amtrak, for some proposed actions, such as increasing 
service to daily frequency, reaching agreement with freight railroads 
is not difficult because the freight railroads’ infrastructure can 
support additional trains and the host freight railroad may already be 
used to having Amtrak operate along certain routes. In other cases—
such as where substantial capital improvements are needed or where 
service is to be initiated over tracks that are operating at or near 
capacity, reaching an agreement might be more difficult, especially 
where Amtrak expects freight railroads to pay for some or all of the 
improvements. 

Amtrak officials told us that they met with senior freight railroad 
officials in November and December 1999-—before its board of directors 
approved the Network Growth Strategy-—to tell them of Amtrak’s plan 
for expanded service. Amtrak officials stated that freight railroads 
did not then express opposition to proposed expanded routes and 
services. According to Amtrak, these were high-level discussions at 
the president/chief executive officer level, during which the railroad 
executives agreed to entertain more specific proposals. According to 
Amtrak, it met again with officials from each railroad, generally in 
January or February 2000, to outline specific route proposals. 
According to an Amtrak official, Amtrak discussed the proposed route 
and/or service actions and sought freight railroads’ overall reaction 
to the proposals. He said that, in some cases, freight railroads 
identified issues such as the need to upgrade track. However, 
generally freight railroads said that they needed to further analyze 
the proposals to determine their likely effect, with more detailed 
discussions to be held at later dates. Freight railroad officials told 
us that the initial and subsequent meetings focused primarily on the 
concept of providing new services rather than identifying whether 
there might be aspects of the proposals that would be easy or 
difficult to resolve.[Footnote 11] 

While Amtrak recognized that capital improvements would be needed on 
freight railroads’ tracks to implement eight Network Growth Strategy 
routes, it did not include capital investment requirements or the 
source of these funds in its route evaluations until after it had 
decided to implement the action.[Footnote 12] An Amtrak official said 
that considering capital investment requirements any earlier would not 
be useful since, if capital costs were factored in, route proposals 
would appear to be unprofitable and not be considered further. As a 
result, Amtrak limited its analysis to whether revenues are expected 
to exceed operating costs. Amtrak followed this approach despite the 
fact that some route actions cannot be implemented-—and its operating 
losses reduced-—unless capital is available. 

It was not until after Amtrak decided to implement the Network Growth 
Strategy in December 1999 and announced it to Congress that it began 
to develop an understanding of the capital investments needed to 
implement the route and service actions and other implementation 
issues critical to gaining freight railroad agreement. For example, it 
was not until spring 2000 that Amtrak learned from the Union Pacific 
Railroad that it might cost about $40 million to implement the 
Crescent Star (service between Meridian, Mississippi, and Dallas/Fort 
Worth, Texas).[Footnote 13] A Union Pacific official told us that his 
railroad was not willing to share the costs of this investment with 
Amtrak, nor was it willing to help Amtrak finance it over time. He 
said that capital investment had not been discussed with his railroad 
prior to this time. 

Freight railroads were also concerned about having a competitor on 
their tracks. All four of the freight railroads we contacted that 
would have been affected by the Network Growth Strategy generally 
acknowledged Amtrak’s statutory authority to operate mail and express 
business. However, all expressed concern about Amtrak’s becoming a 
competitor for their freight business. This concern was heightened by 
Amtrak’s plans to begin running large numbers of express cars on their 
trains as it expanded its mail and express business. This concern 
contributed to Amtrak’s decision to cancel the Skyline service. A 
Norfolk Southern official said his company did not want Amtrak to 
solicit business on this route that was similar to its own freight 
business. Other freight railroads we contacted were similarly wary of 
Amtrak’s plans to use its route and service expansion to increase 
express business that could potentially compete with their own. 

In addition, Amtrak did not identify potential operational problems 
that could be encountered, such as whether capacity constraints would 
be important. A good illustration is Amtrak’s planned Crescent Star 
service. This service, planned for implementation in summer 2000 over 
Union Pacific Railroad and Norfolk Southern lines, has not yet come to 
fruition. 

According to a Union Pacific official, the company could not reach 
agreement with Amtrak, in part, because the planned routing would have 
worsened congestion on the line. In addition, a Norfolk Southern 
official told us that the rail infrastructure in Meridian would not 
support passenger train switching operations without serious 
interference with freight trains. As a result of these operational 
problems and because of funding problems, the routing of this still-to-
be-implemented service has since shifted to another railroad.[Footnote 
14] The proposed Twilight Limited faced similar problems. According to 
CSX Transportation officials, this service could have encountered 
significant line capacity and scheduling problems west of Albany, New 
York. Finally, a Union Pacific official told us that the Aztec Eagle 
(service from San Antonio to Laredo, Texas) could have created 
capacity problems because it would have utilized Union Pacific’s 
primary route to Mexico.[Footnote 15] 

Amtrak officials agreed that routing of the Crescent Star was shifted 
to another railroad because of disagreements with Union Pacific. An 
Amtrak official said Union Pacific was initially receptive to proposed 
route and service actions but turned negative when plans became more 
specific. Amtrak officials also agreed infrastructure improvements 
were necessary in Meridian, Mississippi, but believed these were not 
insurmountable problems. Amtrak officials also did not believe there 
would be significant problems with the Twilight Limited because the 
proposed service was to replace existing trains in both New York and 
Michigan. 

In other instances, Amtrak was not able to reach agreement with 
freight railroads on compensation for track access, especially for 
trains with additional express traffic. Freight railroads often 
receive additional compensation for handling Amtrak trains over a 
certain length and/or for cars carrying express business. Issues of 
compensation contributed to the cancellation of at least one route 
action—-the Skyline. This route—-establishing service between Chicago 
and New York City via Pennsylvania-—involved Norfolk Southern. Norfolk 
Southern officials said they were willing to work with Amtrak on 
establishing this service and had even reached agreement with Amtrak 
about the operating arrangements for this train. (The train was to be 
handled similarly to a regular freight train, including operating at 
60 miles per hour-—a speed closer to freight train speed.) However, 
Norfolk Southern largely attributed the demise of this route action to 
the inability to reach agreement with Amtrak over the compensation to 
be paid for track access and additional express business. 

Conclusions: 

Amtrak’s Network Growth Strategy has been unsuccessful because it 
overestimated (1) revenues expected from new mail and express service 
and (2) its ability to reach agreement with freight railroads over 
capital funding and other implementation issues. Amtrak said that it 
has improved it revenue estimation process. However, reaching 
agreement with freight railroads will always be a major challenge when 
Amtrak attempts to expand its business in areas that are operating at 
or near capacity, when the expansion appears to pose competition for 
freight railroads, or when freight railroads are expected to make 
capital investments to help implement the routes. We believe that, in 
any future major route and service expansions predicated on improving 
Amtrak’s financial condition, Amtrak’s decisionmaking process needs to 
more explicitly reveal the risks associated with successful 
implementation. 

Recommendation for Executive Action: 

We recommend that, for any future major route and service proposals, 
the president of Amtrak disclose to Amtrak’s board of directors any 
significant risks that could impair the successful implementation of 
the planned actions and its plans to ameliorate those risks. These 
potential risks include the expected ability to obtain capital funding 
and reach agreement with freight railroads to operate over their 
tracks. 

Agency Comments and Our Evaluation: 

We provided a draft of this report to Amtrak and to the Department of 
Transportation for their review and comment. Amtrak disagreed with the 
conclusions we reached about the benefits that might have been 
achieved through discussing its strategy with its key partners more 
substantively before—rather than after—deciding to expand its 
operations over freight railroad tracks. Amtrak provided its comments 
during a meeting with the acting vice president for government affairs 
and others and in a subsequent letter. (See app. VI.) The department 
generally agreed with the report’s conclusions via an e-mail message. 

In commenting on a draft of this report, Amtrak agreed with our 
presentation of the reasons that it overestimated mail and express 
revenue. Amtrak also stated that a major theme of our report was that 
Amtrak should have delayed in communicating with Congress the route 
and service changes proposed in February 2000 to allow time for 
additional analysis and negotiations with freight railroads. By so 
doing Amtrak would have proposed considerably fewer new services and 
would have been more successful in implementing its proposals. We are 
not suggesting that Amtrak should have delayed announcing the Network 
Growth Strategy. Rather, our work clearly illustrates the need for 
Amtrak to perform due diligence to understand the likely positions of 
key stakeholders—-whose cooperation is essential to successful route 
and service expansion-—before, rather than after, committing itself to 
implementing them. However, we believe that Amtrak’s not examining 
more closely the capital improvements needed to implement their route 
proposals and whether freight railroads would likely agree to them 
were significant flaws in Amtrak’s strategy. We agree with Amtrak’s 
characterization of our opinion that, if it had a better understanding 
of the concerns of key stakeholders, it might not have proposed all of 
the resulting route actions. We would have viewed a decision to 
implement fewer or different route actions each with a greater 
likelihood of being successfully implemented, rather than a larger 
number of speculative proposals, as sound business judgment because it 
would have increased the likelihood that Amtrak could have realized 
operating profits and moved closer to the goal of reaching operational 
self-sufficiency. During our work, we received conflicting information 
about Amtrak’s early interaction with freight railroads. As a result 
of our meeting with Amtrak, we discussed this topic again with freight 
railroads and Amtrak and revised this report to better show this early 
interaction. 

Amtrak also stated that (1) it needed to act quickly to reach 
operational self-sufficiency within 3 years, (2) the purpose of the 
Network Growth Strategy was to implement route and service changes 
that would more than cover their operating costs and therefore 
contribute to achieving operational self-sufficiency, and (3) not 
every route and service change requires lengthy negotiations. 
Regarding Amtrak’s first point, we agree that there was immense 
pressure on Amtrak to become operationally self-sufficient. However, 
we believe that this pressure made it even more important for Amtrak 
to conduct the due diligence needed before it decided to move ahead. 
Without an understanding of the likelihood that freight railroads 
would be receptive to Amtrak’s plans and that Amtrak could find the 
capital funds needed to implement these changes, Amtrak had little 
basis to expect that the route and service proposals it made could 
actually be implemented expeditiously so as to help reduce Amtrak’s 
need for federal operating subsidies. Amtrak appeared to tacitly 
acknowledge the necessity of doing so, at least where capital funding 
is an issue, when it stated in its comments: “[T]he growing capacity 
constraints on many key lines mean that freight railroads can, not 
infrequently, demand large infusions of capital from passenger train 
operators to accommodate additional trains.” Regarding Amtrak’s second 
point, we agree that Amtrak’s goal was to implement routes in which 
revenues exceeded operating costs. It was not our intention to suggest 
that Amtrak should have only decided to implement routes that covered 
their capital costs too. We have revised our recommendation to remove 
such an impression. Regarding Amtrak’s third point, we agree that some 
proposed route and service changes may be implemented easily and have 
revised our recommendation to more explicitly recognize this condition. 

In our meeting with Amtrak officials, Amtrak disagreed with the 
statement in our draft report that it had poor information on 
interconnectivity (revenues from passengers taking more than one train 
to reach their final destinations). Although this comment conflicts 
with statements made by Amtrak during our work, we acknowledge that 
Amtrak did have data on interconnectivity at the time it was 
performing its market-based network analysis. Accordingly, we have 
deleted references to interconnectivity in this report. 

Finally, Amtrak believes that we did not sufficiently recognize the 
market-based analysis framework was a significant step forward in 
Amtrak’s ability to analyze the market potential for its services. We 
agree that the market-based approach was a significant step forward 
for Amtrak. However, the approach’s usefulness was ultimately 
undermined by Amtrak’s reliance on speculative data on expected 
express business and unrealized assumptions that the route and service 
changes could be implemented quickly and easily. We have added 
information to this report to better portray the differences between 
the market-based analysis framework and Amtrak’s previous approach. We 
also made a number of other revisions throughout this report to better 
portray the extent of Amtrak’s interactions with freight railroads and 
where limits to the interaction led to implementation problems. We 
also made changes, where appropriate, to this report based on our 
meeting with Amtrak. 

The associate administrator for railroad development at the Federal 
Railroad Administration within the Department of Transportation stated 
that the department agreed that Amtrak needs better information on 
which to base its route and service actions. In particular, the 
department agreed with our fundamental conclusions that (1) Amtrak 
needs to undertake earlier negotiation over access-related issues for 
new services and (2) until recently, Amtrak senior management 
incorrectly assumed that it had credible information on mail and 
express revenues and, in particular, costs. 

Scope and Methodology: 

Our work focused on route and service actions that Amtrak considered 
under its market-based approach and Network Growth Strategy. To 
understand the market-based approach and the Network Growth Strategy, 
including its approach to estimating mail and express revenues and 
collaborating with freight railroads, we reviewed documents describing 
the market-based approach, how it works, and the models used for 
financial evaluation. We also reviewed studies done by others to 
identify potential limitations to the market-based approach and 
discussed these limitations with Amtrak and Department of 
Transportation officials. We did not independently evaluate the market-
based approach or its models. As part of our work, we identified route 
and service actions Amtrak has taken since 1995 and the current status 
of the Network Growth Strategy. Finally, we discussed Network Growth 
Strategy route and service actions with officials from Amtrak, four 
major railroads that would have been affected had the Network Growth 
Strategy been fully implemented (the Burlington Northern and Santa Fe 
Railway Company; CSX Transportation, Inc.; the Norfolk Southern 
Corporation; and the Union Pacific Railroad Company), and the state of 
Florida. We conducted our work from July 2001 to April 2002 in 
accordance with generally accepted government auditing standards. 

As arranged with your office, unless you publicly announce its 
comments earlier, we plan no further distribution of this report until 
30 days after the date of this letter. At that time, we will send 
copies of this report to congressional committees with 
responsibilities for intercity passenger rail issues; the acting 
president of Amtrak; the secretary of transportation; the 
administrator, Federal Railroad Administration; and the director, 
Office of Management and Budget. We will also make copies available to 
others upon request. This report will also be available on our home 
page at [hyperlink, http://www.gao.gov]. 

If you or your staff have any questions about this report, please 
contact either James Ratzenberger at ratzenbergerj@gao.gov or me at 
heckerj@gao.gov. Alternatively, we may be reached at (202) 512-2834. 
Key contributors to this report were Helen Desaulniers, Richard 
Jorgenson, Sara Moessbauer, James Ratzenberger, and Edward Warner. 

Sincerely yours, 

Signed by: 

JayEtta Z. Hecker 
Director, Physical Infrastructure Issues: 

[End of section] 

Appendix I: Financial Performance of Amtrak’s Routes, Fiscal Year 2001 

The financial performance of Amtrak routes ranged from an operating 
profit of $51.3 million on the Metroliner/Acela Express to an 
operating loss of $71.5 million on the Northeast Direct. 

Table 2: Operating Profit (Loss), Operating Ratio, and Profit (Loss) 
per Passenger of Each Amtrak Route, Fiscal Year 2001, Ranked by Profit 
(Loss): 

Route: Metroliner/Acela Express; 
State support: $0; 
Total revenue[A]: $271.2 million; 
Total cost[B]: $220.0 million; 
Profit (loss): $51.3 million; 
Operating ratio[C]: 0.8%; 
Profit (loss) per passenger: $19. 

Route: Heartland Flyer; 
State support: $4.6 million; 
Total revenue[A]: $5.8 million; 
Total cost[B]: $5.2 million; 
Profit (loss): $0.6 million; 
Operating ratio[C]: 0.9%; 
Profit (loss) per passenger: $10. 

Route: Vermonter; 
State support: $1.5 million; 
Total revenue[A]: $5.8 million; 
Total cost[B]: $6.4 million; 
Profit (loss): -$0.6 million; 
Operating ratio[C]: 1.1%; 
Profit (loss) per passenger: -$9. 

Route: Adirondack; 
State support: $2.7 million; 
Total revenue[A]: $7.1 million; 
Total cost[B]: $7.8 million; 
Profit (loss): -$0.7 million; 
Operating ratio[C]: 1.1%; 
Profit (loss) per passenger: -$8. 

Route: Piedmont; 
State support: $3.3 million; 
Total revenue[A]: $4.0 million; 
Total cost[B]: $5.0 million; 
Profit (loss): -$1.0 million; 
Operating ratio[C]: 1.3%; 
Profit (loss) per passenger: -$20. 

Route: Lake Country Limited; 
State support: $0; 
Total revenue[A]: $0.1 million; 
Total cost[B]: $2.0 million; 
Profit (loss): -$1.9 million; 
Operating ratio[C]: 36.9%; 
Profit (loss) per passenger: -$1,208. 

Route: Kansas City-St. Louis; 
State support: $6.1 million; 
Total revenue[A]: $10.5 million; 
Total cost[B]: $12.6 million; 
Profit (loss): -$2.1 million; 
Operating ratio[C]: 1.2%; 
Profit (loss) per passenger: -$12. 

Route: Ethan Allen Express; 
State support: $0.2 million; 
Total revenue[A]: $2.2 million; 
Total cost[B]: $4.5 million; 
Profit (loss): -$2.2 million; 
Operating ratio[C]: 2.0%; 
Profit (loss) per passenger: -$53. 

Route: Pere Marquette; 
State support: $2.2 million; 
Total revenue[A]: $4.1 million; 
Total cost[B]: $6.6 million; 
Profit (loss): -$2.5 million; 
Operating ratio[C]: 1.6%; 
Profit (loss) per passenger: -$43. 

Route: Illinois Zephyr; 
State support: $2.8 million; 
Total revenue[A]: $5.5 million; 
Total cost[B]: $8.2 million; 
Profit (loss): -$2.7 million; 
Operating ratio[C]: 1.5%; 
Profit (loss) per passenger: -$27. 

Route: International; 
State support: $3.7 million; 
Total revenue[A]: $7.1 million; 
Total cost[B]: $10.0 million; 
Profit (loss): -$2.9 million; 
Operating ratio[C]: 1.4%; 
Profit (loss) per passenger: -$27. 

Route: Illini; 
State support: $2.4 million; 
Total revenue[A]: $6.0 million; 
Total cost[B]: $9.1 million; 
Profit (loss): -$3.1 million; 
Operating ratio[C]: 1.5%; 
Profit (loss) per passenger: -$30. 

Route: Carolinian; 
State support: $2.7 million; 
Total revenue[A]: $16.2 million; 
Total cost[B]: $20.2 million; 
Profit (loss): -$4.0 million; 
Operating ratio[C]: 1.1%; 
Profit (loss) per passenger: -$16. 

Route: Capitols; 
State support: $18.4 million; 
Total revenue[A]: $30.2 million; 
Total cost[B]: $34.6 million; 
Profit (loss): -$4.4 million; 
Operating ratio[C]: 1.2%; 
Profit (loss) per passenger: -$4. 

Route: Kentucky Cardinal; 
State support: $0; 
Total revenue[A]: $1.4 million; 
Total cost[B]: $7.6 million; 
Profit (loss): -$6.2 million; 
Operating ratio[C]: 5.4%; 
Profit (loss) per passenger: -$212. 

Route: Cascades; 
State support: $16.3 million; 
Total revenue[A]: $31.8 million; 
Total cost[B]: $38.1 million; 
Profit (loss): -$6.3 million; 
Operating ratio[C]: 1.2%; 
Profit (loss) per passenger: -$11. 

Route: San Joaquins; 
State support: $19.5 million; 
Total revenue[A]: $43.0 million; 
Total cost[B]: $52.0 million; 
Profit (loss): -$9.0 million; 
Operating ratio[C]: 1.2%; 
Profit (loss) per passenger: -$13. 

Route: Auto Train; 
State support: $0; 
Total revenue[A]: $54.6 million; 
Total cost[B]: $66.4 million; 
Profit (loss): -$11.8 million; 
Operating ratio[C]: 1.2%; 
Profit (loss) per passenger: -$55. 

Route: Cardinal; 
State support: $0; 
Total revenue[A]: $4.4 million; 
Total cost[B]: $17.1 million; 
Profit (loss): -$12.6 million; 
Operating ratio[C]: 3.9%; 
Profit (loss) per passenger: -$187. 

Route: Hiawathas; 
State support: $5.1 million; 
Total revenue[A]: $12.6 million; 
Total cost[B]: $26.0 million; 
Profit (loss): -$13.3 million; 
Operating ratio[C]: 2.1%; 
Profit (loss) per passenger: -$31. 

Route: Chicago-St. Louis; 
State support: $3.8 million; 
Total revenue[A]: $11.5 million; 
Total cost[B]: $27.7 million; 
Profit (loss): -$16.1 million; 
Operating ratio[C]: 2.4%; 
Profit (loss) per passenger: -$64. 

Route: Keystone and Clocker service; 
State support: $2.8 million; 
Total revenue[A]: $45.2 million; 
Total cost[B]: $65.6 million; 
Profit (loss): -$20.4 million; 
Operating ratio[C]: 1.5%; 
Profit (loss) per passenger: -$7. 

Route: Silver Meteor; 
State support: $0; 
Total revenue[A]: $28.5 million; 
Total cost[B]: $49.8 million; 
Profit (loss): -$21.2 million; 
Operating ratio[C]: 1.7%; 
Profit (loss) per passenger: -$84. 

Route: Chicago-Pontiac; 
State support: $0; 
Total revenue[A]: $9.7 million; 
Total cost[B]: $30.9 million; 
Profit (loss): -$21.2 million; 
Operating ratio[C]: 3.2%; 
Profit (loss) per passenger: -$72. 

Route: City of New Orleans; 
State support: $0; 
Total revenue[A]: $15.3 million; 
Total cost[B]: $39.1 million; 
Profit (loss): -$23.7 million; 
Operating ratio[C]: 2.6%; 
Profit (loss) per passenger: -$127. 

Route: Capitol Limited; 
State support: $0; 
Total revenue[A]: $21.4 million; 
Total cost[B]: $45.6 million; 
Profit (loss): -$24.2 million; 
Operating ratio[C]: 2.1%; 
Profit (loss) per passenger: -$157. 

Route: Pacific Surfliner; 
State support: $21.3 million; 
Total revenue[A]: $52.5 million; 
Total cost[B]: $78.6 million; 
Profit (loss): -$26.1 million; 
Operating ratio[C]: 1.5%; 
Profit (loss) per passenger: -$15. 

Route: Pennsylvanian; 
State support: $0; 
Total revenue[A]: $9.2 million; 
Total cost[B]: $35.4 million; 
Profit (loss): -$26.3 million; 
Operating ratio[C]: 3.9%; 
Profit (loss) per passenger: -$292. 

Route: Silver Palm; 
State support: $0; 
Total revenue[A]: $28.3 million; 
Total cost[B]: $57.0 million; 
Profit (loss): -$28.7 million; 
Operating ratio[C]: 2.0%; 
Profit (loss) per passenger: -$131. 

Route: Silver Star; 
State support: $0; 
Total revenue[A]: $30.7 million; 
Total cost[B]: $60.8 million; 
Profit (loss): -$30.0 million; 
Operating ratio[C]: 2.0%; 
Profit (loss) per passenger: -$113. 

Route: Three Rivers; 
State support: $0; 
Total revenue[A]: $26.5 million; 
Total cost[B]: $59.3 million; 
Profit (loss): -$32.8 million; 
Operating ratio[C]: 2.2%; 
Profit (loss) per passenger: -$245. 

Route: Crescent; 
State support: $0; 
Total revenue[A]: $30.8 million; 
Total cost[B]: $65.8 million; 
Profit (loss): -$35.0 million; 
Operating ratio[C]: 2.1%; 
Profit (loss) per passenger: -$132. 

Route: Empire Service/Maple Leaf; 
State support: $0; 
Total revenue[A]: $52.5 million; 
Total cost[B]: $89.0 million; 
Profit (loss): -$36.5 million; 
Operating ratio[C]: 1.7%; 
Profit (loss) per passenger: -$28. 

Route: Sunset Limited; 
State support: $0; 
Total revenue[A]: $17.7 million; 
Total cost[B]: $56.1 million; 
Profit (loss): -$38.3 million; 
Operating ratio[C]: 3.2%; 
Profit (loss) per passenger: -$347. 

Route: Texas Eagle; 
State support: $0; 
Total revenue[A]: $22.4 million; 
Total cost[B]: $60.7 million; 
Profit (loss): -$38.4 million; 
Operating ratio[C]: 2.7%; 
Profit (loss) per passenger: -$258. 

Route: Lake Shore Limited; 
State support: $0; 
Total revenue[A]: $30.6 million; 
Total cost[B]: $72.4 million; 
Profit (loss): -$41.9 million; 
Operating ratio[C]: 2.4%;
Profit (loss) per passenger: -$143. 

Route: Empire Builder; 
State support: $0; 
Total revenue[A]: $53.3 million; 
Total cost[B]: $98.7 million; 
Profit (loss): -$45.4 million; 
Operating ratio[C]: 1.9%;
Profit (loss) per passenger: -$114. 

Route: Coast Starlight; 
State support: $0; 
Total revenue[A]: $41.2 million; 
Total cost[B]: $87.1 million; 
Profit (loss): -$45.9 million; 
Operating ratio[C]: 2.1%;
Profit (loss) per passenger: -$93. 

Route: California Zephyr; 
State support: $0; 
Total revenue[A]: $51.7 million; 
Total cost[B]: $103.7 million; 
Profit (loss): -$52.0 million; 
Operating ratio[C]: 2.0%; 
Profit (loss) per passenger: -$144. 

Route: Southwest Chief; 
State support: $0; 
Total revenue[A]: $65.9 million; 
Total cost[B]: $128.7 million; 
Profit (loss): -$62.8 million; 
Operating ratio[C]: 2.0%; 
Profit (loss) per passenger: -$237. 

Route: Northeast Direct/Acela Regional/Twilight Shoreliner; 
State support: $0; 
Total revenue[A]: $328.6 million; 
Total cost[B]: $400.1 million; 
Profit (loss): -$71.5 million; 
Operating ratio[C]: 1.2%; 
Profit (loss) per passenger: -$11. 

Route: Total/average; 
State support: $119.2 
Total revenue[A]: $1.497 billion; 
Total cost[B]: $2.271 billion; 
Profit (loss): -$774.2 million; 
Operating ratio[C]: 1.5%[D]; 
Profit (loss) per passenger: -$33.0 million[D]. 

Notes: 

Includes only the revenue and expenses associated with providing core 
intercity passenger rail service along the route. These core services 
are passenger-related services, mail and express, other transportation 
services, and states’ payments supporting certain routes. Excludes 
$4.3 million in unallocated train labor costs, which, according to 
Amtrak, are not clearly identified with specific routes but are 
incurred to provide train service. Excludes depreciation in order to 
show the cash results by route. Does not include special trains. All 
dollar figures are rounded to the nearest hundred thousand dollars, 
except profit (loss) per passenger, which is rounded to the nearest 
dollar. 

Several of the amounts, including the overall loss, differ from those 
that Amtrak reported to Congress. Amtrak included profits from non-
core business (such as commuter contract services) in the results that 
it reported to Congress. We have excluded these amounts because such 
costs are unrelated to route financial performance. 

[A] Includes state support. 

[B] Includes train, route, and system costs. Train costs include such 
factors as crew salaries, fuel and power costs, and payments to 
freight railroads for the use of their tracks. Route costs include 
such factors as maintenance for Amtrak-owned stations. System costs 
include such factors as staff salaries, rent, and associated expenses 
for corporate and strategic business unit headquarters operations. 

[C] Operating ratio is total expenses divided by total revenues. 

[D] Represents a weighted average. 

Source: GAO’s analysis of Amtrak’s data. 

[End of table] 

[End of section] 

Appendix II: Amtrak Route Actions, January 1995 Through December 2001: 

Year: 1995: 

Route: Cardinal; 
Action: Truncated, 1995. 

Route: Broadway Limited; 
Action: Eliminated, 1995. 

Route: Atlantic City Express; 
Action: Eliminated, 1995. 

Route: Palmetto; 
Action: Eliminated, 1995. 

Route: Hoosier State; 
Action: Frequency reduced, 1995; eliminated, 1995. 

Route: Lake Cities; 
Action: Truncated, 1995. 

Route: Texas Eagle; 
Action: Truncated, 1995, 1997; extended 1998; frequency increased, 
1998, 2000; frequency decreased, 2001. 

Route: Gulf Breeze; 
Action: Frequency reduced, 1995; eliminated, 1995. 

Route: Desert Wind; 
Action: Frequency reduced, 1995; eliminated, 1997. 

Route: Empire Builder; 
Action: Frequency reduced, 1995; frequency restored, 1997. 

Route: Crescent; 
Action: Frequency reduced, 1995; frequency restored, 1996. 

Route: City of New Orleans; 
Action: Frequency reduced, 1995; frequency restored, 1997. 

Route: California Zephyr; 
Action: Frequency reduced, 1995; frequency restored, 1997. 

Route: Three Rivers; 
Action: Launched, 1995. 

Route: Metroliners[A]; 
Action: Frequency reduced, 1995, 1996. 

Route: Northeast Direct[A]; 
Action: Frequency reduced, 1995. 

Route: Empire[A](New York City-Albany); 
Action: Frequency reduced, 1995; frequency increased, 1998, 1999. 

Route: Empire[A] (Albany-Buffalo); 
Action: Frequency reduced, 1995; frequency increased, 1998. 

Route: Boston-Springfield, Massachusetts[C]; 
Action: Eliminated, 1995; restored, 1996. 

Route: Hiawatha; 
Action: Frequency reduced, 1995. 

Route: New York City-Newport News, Virginia; 
Action: Frequency reduced, 1995, eliminated after 1997. 

Route: Pere Marquette; 
Action: Frequency reduced, 1995; restored in 1996. 

Route: Mule (St. Louis-Kansas City); 
Action: Eliminated, 1995; restored later in 1995. 

Route: Vermonter; 
Action: Launched, 1995. 

Route: Montrealer; 
Action: Eliminated, 1995. 

Route: Niagara Rainbow; 
Action: Eliminated, 1995. 

Route: Keystone[A]; 
Action: Frequency reduced, 1995. 

Route: The Loop; 
Action: Frequency reduced, 1995, eliminated, 1996. 

Route: San Diegan; 
Action: Frequency reduced, 1995; frequency restored, 1997; frequency 
increased, 1998. 

Route: Mount Baker International; 
Action: Launched, 1995. 

Route: Piedmont; 
Action: Launched, 1995. 

Year: 1996: 

Route: Sunset Limited; 
Action: Truncated, 1996, extended, 1997. 

Route: Ethan Allen; 
Action: Launched, 1996, extended, 1998. 

Route: Silver Palm; 
Action: Launched, 1996. 

Route: Gulf Coast Limited; 
Action: Launched, 1996; eliminated, 1997. 

Year: 1997: 

Route: Pioneer; 
Action: Eliminated, 1997. 

Route: Night Owl; 
Action: Eliminated, 1997. 

Route: Detroit-Pontiac; 
Action: Frequency reduced, 1997. 

Route: Twilight Shoreliner[A]; 
Action: Launched, 1997. 

Year: 1998: 

Route: Capitols; 
Action: Frequency increased, 1998, frequency increased, 1999, 2000, 
2001; truncated, 2000. 

Route: Pennsylvanian; 
Action: Launched, 1998. 

Route: Pacific Northwest Corridor[B]; 
Action: Frequency increased, 1998. 

Year: 1999: 

Route: Kentucky Cardinal; 
Action: Launched, 1999; extended, 2001. 

Route: Heartland Flyer; 
Action: Launched, 1999. 

Route: San Jaoquin; 
Action: Frequency increased, 1999. 

Route: Cascades; 
Action: Frequency increased, 1999, 2000. 

Year: 2000: 

Route: Pacific Surfliner; 
Action: Frequency increased, 2000, 2001. 

Route: Lake Country Limited; 
Action: Launched, 2000; eliminated, 2001. 

Route: Acela (regional/express); 
Action: Launched, 2000; frequency increased, 2001. 

Year: 2001: 

Route: Downeaster; 
Action: Launched, 2001. 

[A] Now part of Acela regional service. 

[B] Now part of Cascades service. 

[C] This service is shown separately so as to not distort route and 
service actions affecting the Boston-Washington, D.C., spine of the 
Northeast Direct route. 

Source: Amtrak. 

[End of table] 

[End of section] 

Appendix III: Planned Route and Service Actions Included in the 
Network Growth Strategy: 

The following are the planned route and service actions included in 
the Network Growth Strategy announced by Amtrak in February 2000. 

Table 3: Planned Network Growth Strategy Route and Service Actions: 

Name: Hiawatha; 
Description: Extend Hiawatha service from Chicago-Milwaukee, 
Wisconsin, to Fond du Lac, Wisconsin. 

Name: Lake Cities; 
Description: Extend Chicago-Michigan service to Toledo, Ohio. 

Name: Lake Country Limited; 
Description: New service between Chicago and Janesville, Wisconsin. 
Skyline (Manhattan Limited) New service between New York City and 
Chicago via Philadelphia and Cleveland. 

Name: Silver Meteor; 
Description: Extend current service from New York City to Boston. 

Name: Crescent Star; 
Description: Split Crescent service in Meridian, Mississippi, to 
provide new service from Meridian to Dallas/Ft. Worth. 

Name: Aztec Eagle; 
Description: New service between San Antonio, Texas, and Monterrey, 
Mexico. 

Name: Texas Eagle; 
Description: Increase service to daily from 4 times per week. 

Name: Twilight Limited; 
Description: New service between New York City and Chicago through 
upstate New York and Canada. 

Name: Luxury Transcontinental; 
Description: New luxury transcontinental service (origin and 
destination to be determined). 

Name: International; 
Description: Reroute existing train service between Battle Creek, 
Michigan, and Toronto, Canada. 

Name: Hawkeye; 
Description: New service between Chicago and Des Moines, Iowa. 

Name: Silver Service; 
Description: Various reroutes of Amtrak Silver service trains in 
Florida, including new service from Jacksonville to Miami via the 
Florida East Coast Railway. 

Name: Sunset Limited; 
Description: Reroute train in Texas from El Paso to San Antonio to El 
Paso to Ft. Worth via Abilene. 

Name: Kentucky Cardinal[A]; 
Description: Extend service from Jeffersonville, Indiana, to 
Louisville, Kentucky. 

[A] The extension of the Kentucky Cardinal to Louisville was 
implemented in December 2001. Source: GAO’s analysis of Amtrak’s data. 

Source: GAO's analysis of Amtak's data. 

[End of table] 

[End of section] 

Appendix IV: Amtrak’s Process for Evaluating Route and Service 
Proposals: 

Amtrak officials told us that route and service decisions primarily 
focus on whether the increased operating revenues from an action are 
expected to exceed the direct costs. Increased revenues can arise from 
adding passengers on the route, adding passengers on the route who can 
then transfer to other routes and vice versa (called 
interconnectivity) and from hauling mail and express. While Amtrak 
recognizes early in its planning process that it will incur costs for 
capital expenditures (e.g., to acquire equipment and facilities and to 
finance needed track and signal improvements) and, to a lesser extent, 
state financial support, it typically does not address these needs 
until after it has decided to implement a route or service action. 

Amtrak’s Consideration of Operating Revenue and Direct Costs: 

Ideas for route and service actions are either generated internally or 
from those outside Amtrak seeking additional service. To provide a 
quick assessment of an idea’s reasonability, Amtrak informally 
examines issues such as the number and type of equipment (e.g., cars 
and locomotives) that might be needed, where the train might stop, and 
possible train schedules. If a proposal appears promising, Amtrak 
begins a more formal evaluation process. First, it estimates potential 
ridership, from which it derives passenger-related revenue estimates. 
To do so, Amtrak uses one of two models, depending on whether the 
action involves a long-distance route or a shorter-distance route 
serving a transportation corridor (such as that between Washington, 
D.C., and New York City). Of the two models, only the model for 
transportation corridors can assess the potential market share that a 
proposed route action will attract from among a corridor’s various 
transportation modes. For example, this model can estimate the impact 
of a shift from rail ridership to automobile usage prompted by a 
decline in gas prices. The ridership projections are, in turn, used to 
estimate a route’s passenger-related unit costs and operating costs. 
An Amtrak official said that cost estimates often increase later in 
the evaluation process as more amenities are added to a proposed 
service as a means to attract more riders. 

After it completes its initial assessments, Amtrak uses its market-
based approach to model the expected financial impact of the route or 
service action. It models the proposed route and service action 
individually and as part of the whole route network. In some cases, 
Amtrak will model several variations of a proposed route action to see 
if one is more financially viable than others. 

Amtrak also estimates potential mail and express revenues associated 
with proposed route actions. Mail revenue estimates are largely based 
on contracts Amtrak has with the U.S. Postal Service, discussions with 
U.S. Postal Service officials, and U.S. Postal Service projections. 
Estimating express revenue is somewhat more difficult. According to 
Amtrak, until recently, it estimated express revenue largely on the 
basis of an analysis of a database of commodities being shipped 
nationally. Amtrak estimated the portion of this business that it 
thought it could obtain. An Amtrak official said that it now focuses 
more on determining existing customers’ shipping needs, assessing 
these needs in light of current economic trends, and evaluating 
Amtrak’s ability to meet these needs given existing train capacity. 
Amtrak relies on its mail and express unit to estimate mail and 
express costs. However, Amtrak officials told us identifying these 
costs has been difficult since Amtrak did not formerly identify these 
costs separately but rather incorporated them into other business 
units. Amtrak currently has a project under way to identify the 
specific costs of its mail and express business. 

When Amtrak management wants to proceed with a route action, it either 
seeks the approval of its board of directors or directs the affected 
business unit to implement the action.[Footnote 16] Amtrak policy 
requires board approval to initiate service on new routes or to 
discontinue service on routes. Amtrak said its strategic business 
units have the authority to make minor changes in the schedules and 
frequencies of their train service. 

Consideration of Capital Costs and Other Financial Issues: 

Amtrak officials told us that the company also considers the cost of 
capital improvements that may be associated with route actions, the 
fees that freight railroads will charge for access to their tracks, 
and the likelihood states might be interested in financially 
supporting the routes. While Amtrak told us that it recognizes early 
in its planning process that capital costs may be incurred for routes 
other than its Northeast Corridor, it does not formally consider these 
costs under the market-based approach until after it decides to 
implement a route or service change. Amtrak officials said that they 
do not consider these costs earlier because Amtrak’s lack of 
investment capital would preclude further consideration of such 
proposals. Amtrak’s capital funds most often come from federal 
assistance and from freight railroads and states that might agree to 
contribute funds. 

Amtrak estimates track access payments based on operating agreements 
with freight railroads, or in the case of railroads it has not dealt 
with before, it is assumed access fees will not vary significantly 
from national averages. Obtaining state financial support for routes 
generally varies depending on the length of the route. An Amtrak 
official said the corporation aggressively pursues state support on 
short-distance, commuter-like routes. In fact, he said states often 
approach Amtrak about initiating or expanding this type of service on 
their own. However, on longer distance routes that go through many 
states, most states are not interested in providing financial support 
and Amtrak must assume the financial responsibility itself. In 2001, 
states provided financial support for 18 of Amtrak’s 43 routes. This 
support ranged from about $200,000 on the Ethan Allen Express route 
(service between New York and Vermont) to about $21 million on the 
Pacific Surfliner (service within California). 

[End of section] 

Appendix V: Market-Based Network Analysis Models Used to Estimate 
Ridership, Revenues, and Costs: 

The market-based framework includes a set of models used to predict 
changes in ridership, revenues, and costs likely to result from a 
planned restructuring of Amtrak’s route system or a variation in 
service levels on existing routes. A series of demand models estimate 
ridership and passenger-related revenues in response to variation in 
the stations served by each route, train departure frequencies or 
schedules, travel times, and fare structures. Then a series of 
financial models translate operating statistics that are based on the 
type of equipment and level of operations (e.g., level of onboard 
staffing) required into operating expenses. For high-speed rail 
scenarios, a capital cost model estimates the capital that is likely 
to be required (e.g., upgrade to track and additional rolling stock) 
to make changes in service.[Footnote 17] 

Models Used to Estimate Ridership and Revenue: 

Amtrak uses separate models to estimate ridership and revenue for 
short-distance intercity routes (regional networks where frequent, 
higher-speed services are planned, called “corridors”) and long-
distance intercity routes (typically at least 500 miles long). For 
short-distance intercity routes, the market-based approach includes a 
model specifically for the Northeast Corridor (Boston-Washington), as 
well as a generic model to predict ridership and revenue in corridors 
located in other regions. Each of the corridor models is designed to 
be a mode-split model. That is, these models predict the share of 
travel likely to take place by automobile, air, and rail, based on the 
projected level of total traffic for each market. Amtrak typically 
estimates total traffic between two areas on the basis of expected 
demographic and economic growth in the areas making up the corridor. 
Then it estimates the market shares for each mode on the basis of the 
costs of traveling by each mode and the level of service it provides 
in the corridor (departure frequency and total travel time). Finally, 
it uses the estimated ridership with the assumed fare structures to 
estimate the passenger-related revenue that would result from a 
proposed route. 

The long-distance demand model consists of two components. The first 
component predicts the total number of rail passengers traveling 
between each station-pair and the second component predicts the class 
of service (mainly sleeper versus coach). The ridership estimate is 
based on such factors as population and employment in the surrounding 
areas. Unlike the corridor demand models, this is not a mode-split 
model, but rather a direct rail model. As such, the model does not 
directly assess the amount of total traffic in the market or the 
amount of traffic that may be captured by alternative modes. Instead, 
it employs historic data from a sample of markets to assess the 
relationships between rail ridership levels and such factors as 
population, employment, travel time, and level of rail service. Amtrak 
then uses these estimated associations to estimate the level of rail 
traffic in a new market. After projecting the rail ridership, Amtrak 
uses the second component to estimate what fraction of passengers will 
choose each of four classes of service, based on such factors as 
frequency and timing of trains and the fares charged for each service 
level. Finally, Amtrak uses the projected levels of ridership and 
class of service to estimate the passenger-related revenue that would 
result from a proposed route. 

Models Used to Estimate Costs: 

The market-based approach includes operating and, in the case of high-
speed rail scenarios, capital cost models, to estimate the likely 
impact on Amtrak’s expenses from planned changes to its network. These 
models translate operating statistics into operating or “direct” 
expenses. The operating statistics are developed based on ridership 
patterns, schedule, train makeup, and staffing data for an individual 
route into estimates of the equipment fleet and train crew 
requirements, as well as the number of monthly train-miles. The 
operating model applies detailed unit cost to predict the changes in 
operating expenses for the route. 

The capital cost model estimates the capital investments necessary to 
upgrade existing track or construct new rights-of-way for routes in 
which Amtrak is considering improving travel times, increasing 
frequency, or introducing new services. The model also estimates costs 
for acquiring new rolling stock and other equipment as necessary. The 
requirements for alternative levels of service were developed from 
engineering studies of facility and equipment requirements necessary 
to upgrade a sample of route segments. The model also makes use of 
data on unit cost factors required for upgrading facilities and 
equipment. These data are based on past experience with upgrades in 
several markets. Using the estimated facility and/or equipment needs 
and the data on unit costs, the model calculates an estimate of 
required capital. According to Amtrak, this model is currently applied 
only to high-speed rail corridors. It is not used to determine 
potential capital costs on non-high-speed rail corridors. 

[End of section] 

Appendix VI: Comments from the National Railroad Passenger Corporation: 

Note: GAO comments supplementing those in the report text appear at 
the end of this appendix. 

National Railroad Passenger Corporation: 
60 Massachusetts Avenue, NW: 
Washington, DC 20002: 

March 15, 2002: 

Ms. JayEtta Z. Hecker: 
Director, Physical Infrastructure Issues: 
U.S. General Accounting Office: 
Washington, DC 20548: 

Dear Ms. Hecker: 

Set forth below are Amtrak's comments on GAO Report No. GA0-02-398, 
entitled "Intercity Passenger Rail: Amtrak Needs to Improve Its 
Evaluation of Route and Service Proposals." 

The major theme of this report is that Amtrak should have delayed in 
communicating to Congress the route and service changes it proposed in 
February 2000, as part of its "Network Growth Strategy" 
("NGS"),[Footnote 1] to allow time for additional analyses and 
negotiations with freight railroads. GAO appears to believe that, if 
Amtrak had followed this approach, it would have proposed to add 
considerably fewer new services but would have been more successful in 
implementing its proposals. 

In the past five years, Amtrak and GAO have worked closely on the 
development of over a dozen reports. Amtrak has grown to appreciate 
the work of GAO's staff, particularly their careful attention to 
detail and their receptivity to consultation, and we have generally 
expressed a high level of agreement with GAO's conclusions. In this 
instance, however, Amtrak disagrees with some of the findings and 
observations in the report. 

The Network Growth Strategy: 

The NGS was a proposal for restructuring and expanding Amtrak's route 
system in a manner that would enhance Amtrak's market reach and 
generate additional net revenues. It also represented the first 
comprehensive analysis of Amtrak's national network, and the first 
attempt to apply rigorous financial analyses and modeling techniques 
to its design. 

Before the NGS, decisions regarding changes in Amtrak's route system 
and long-distance train frequencies ("route and service changes") were 
not based upon quantitative analyses of net financial impacts. Rather, 
the addition of new routes typically resulted from Congressional 
directives, and the elimination of routes was constrained by the 
statutory requirement that Amtrak operate a "basic system" of routes 
designated by DOT. 

Not surprisingly, perhaps, when the "basic system" requirement was 
eliminated in 1997, at the same time that Amtrak was given a statutory 
mandate to achieve "operational self-sufficiency" by December 2002, 
Amtrak lacked the tools it needed to analyze the financial impacts of 
route and service changes. As DOT's Inspector General noted in his 
1998 "Independent Assessment of Amtrak's Financial Needs", DOT Report 
No. TR-1999-027, the only available tool for calculating costs -- 
Amtrak's "Route Profitability System" ("RPS") -- "is not appropriate 
for assessing changes in specific routes and services". Because it is 
based upon "fully allocated costs . RPS cannot tell Amtrak which costs 
would change as a result of service changes."[Footnote 2] 

The NGS Report, reviewed by the GAO, was the first product of the 
Market Based Network Analysis ("MBNA") Project that Amtrak initiated 
in October 1998 to address these deficiencies. The timing of the 
issuance of the NGS Report, and the timetable for implementing its 
proposals, was driven by two deadlines. First, route and service 
changes with positive financial impacts would have to be implemented 
quickly if they were to contribute to Amtrak's efforts to meet the 
statutory self-sufficiency deadline. Second, the May 1999 report of 
the Senate Appropriations Committee (Senate Report 106-55) directed 
Amtrak to report the MBNA Project's conclusions regarding route and 
service changes to the Appropriations Committee prior to the issuance 
of Amtrak's FY2000 Strategic Business Plan. [See comment 1] 

In developing the sophisticated MBNA models that were used to 
determine the net impacts of route and service changes, Amtrak drew 
upon extensive data on the costs and revenues associated with the 
operation of its passenger trains, supplemented by survey research to 
quantify passenger demand and "interconnectivity" (the impact of 
changes in one service on connecting services). However, for 
express -- the expedited transportation of premium shipments on 
Amtrak's passenger trains -- reliable historical or cost data was 
simply not available when the MBNA Project was initiated in 1998. 
Amtrak's express initiative had only commenced in the previous year. 
Therefore, the express revenue and cost assumptions utilized in the 
MBNA Project were, by necessity, based upon projections and estimates 
derived from Amtrak's very limited experience. 

Amtrak's Comments on the GAO Report: 

Although the report faults Amtrak for failing to perform more rigorous 
analyses, it barely mentions the MBNA models,[Footnote 3] which are 
discussed only in a two-paragraph appendix on the very last page. 
Instead, the report claims that Amtrak did not adequately consult with 
freight railroads before releasing the NGS Report, and therefore had 
unrealistic expectations as to how they would react. It contends that 
Amtrak should have estimated the capital costs associated with each 
potential route and service change before issuing the NGS Report, and 
factored into its decisions likely capital contributions from specific 
states. The report also states that Amtrak overestimated revenues and 
underestimated costs associated with the carriage of express on new 
routes and services proposed in the NGS Report. 

As discussed in detail below, Amtrak disagrees with a number of the 
report's conclusions regarding two of the principal issues it 
addresses: freight railroad consultations and the projection of 
capital costs. As for express, the report acknowledges that Amtrak no 
longer uses the (best available at the time) methodologies that 
prompted GAO's criticisms. 

I. Consultations with Freight Railroads: 

Before Amtrak issued the NGS Report on February 28, 2000, it met at 
least twice with virtually all of the major freight railroads to 
discuss its NGS proposals, first at the President/CEO level and then 
in a follow-up meeting where Amtrak made customized presentations to 
senior railroad personnel. Nonetheless, the GAO asserts that, before 
adopting the NGS proposals and providing the NGS Report to Congress, 
Amtrak should have engaged in more extensive "consultations" with 
these railroads. 

When the NGS Report was issued, Amtrak faced two pressing deadlines: 
one for delivering the report to Congress, and another for achieving 
operational self-sufficiency (to which the NGS proposals were expected 
to contribute). As discussed below, Amtrak also needed to begin 
developing, as quickly as possible, public support for NGS proposals 
that would require state and local funding. These considerations made 
it necessary for Amtrak to act with expediency. Therefore, while 
advance discussions with freight railroads were appropriate and did 
occur, it was not practical for Amtrak to delay the release of its NGS 
proposals while it engaged in lengthy "consultations" with each of the 
16 affected freight railroads to ensure that it would be able to 
implement each of the 15 proposals. [See comment 3] 

Amtrak does not agree with the GAO's assumption that the 
implementation of route and service changes on freight railroads' 
lines, in every case, requires lengthy negotiations with the host 
freight railroad. Frankly, there is no "model" for how and on what 
timetable such issues are resolved. 

In many cases, Amtrak has been able to implement significant route and 
service changes fairly quickly. For example, one of the NGS-proposed 
changes — increasing the frequency of the St. Louis-to-San Antonio 
"Texas Eagle" from tri-weekly to daily — was implemented on May 27, 
2000, just three months after issuance of the NGS Report, even though 
it required negotiations with two major freight railroads (BNSF and 
UP) regarding additional Amtrak operations on more than 1000 miles of 
their tracks. [See comment 4] The NGS-proposed Chicago-to-Janesville, 
WI service — which involved operations over three different 
railroads — was also implemented in the Spring of 2000. Similarly, 
while the report implies that Amtrak had no reason to expect Norfolk 
Southern to agree to the operation of the Harrisburg-to-Chicago 
segment of the proposed "Skyline (Connection)", Amtrak had implemented 
a similar additional service — an extension of the "Pennsylvanian" — 
over the same route a year before issuance of the NGS, and just eight 
months after notifying Norfolk Southern's predecessor of its desire to 
do so.[Footnote 4] [See comment 5] 

The principal example that the report cites to illustrate its 
assertions that Amtrak did not adequately consult with freight 
railroads involves the proposed Meridian, MS-to-Dallas/Fort Worth 
"Crescent Star". However, the delay in the implementation of that 
train is not due to freight railroad opposition. To the contrary, the 
"Crescent Star" is an example of how Amtrak was able, in cooperation 
with a freight railroad, to develop a "win/win" solution to funding 
capital investment needs. [See comment 6] 

* While the report highlights complaints about the "Crescent Star" by 
Norfolk Southern and Union Pacific, Norfolk Southern's only 
involvement with the "Crescent Star" is at the point in Meridian where 
the new train would diverge from the existing "Crescent" route, and 
Union Pacific owns only a few miles of the train's planned route in 
the Dallas area. The primary railroad involved is the Kansas City 
Southern, which owns the only rail line between Meridian, MS, and 
Shreveport, LA, and one of the two parallel rail lines between 
Shreveport and the Dallas/Fort Worth area. Kansas City Southern, to 
which it appears GAO has not spoken, has strongly supported Amtrak's 
proposal. 

* The NGS Report (p. 21) indicated that, west of Shreveport, the train 
would operate over either Kansas City Southern or Union Pacific, and 
Amtrak approached both railroads. When Union Pacific sought a very 
large capital contribution, Amtrak opted to route the "Crescent Star" 
over Kansas City Southern. We submit that this is not evidence of 
flaws in Amtrak's decisionmaking, but rather a prudent business 
decision. 

* Implementation of the "Crescent Star" has been delayed by 
controversial regulations issued last year that impose prohibitive 
collateral and other requirements on the RRIF loan program with which 
track and signal improvements on the train's route were to be 
financed. Key members of Congress have introduced legislation that 
would overturn these requirements. 

In those instances where Amtrak was unsuccessful in addressing freight 
railroads' concerns about its NGS proposals, the GAO's focus on 
Amtrak's timing and tactics ignores more fundamental issues. The NGS 
may have been the last real attempt to restructure and (modestly) 
expand the rail passenger network in the United States without 
substantial expenditures of federal funds. The growing capacity 
constraints on many key rail lines mean that freight railroads can, 
not infrequently, demand large infusions of capital from passenger 
train operators to accommodate additional trains. 

In the upcoming debate over Amtrak's reauthorization and proposals for 
the "privatization" of intercity rail passenger service, one of the 
most critical issues will be the extent to which federal funding is 
required to support intercity rail passenger service. Some who propose 
changes in Amtrak's responsibilities and structure have asserted that 
the need for federal funding has been overstated, and that states, 
freight railroads, and other parties would be prepared in the near 
term to provide significant additional capital funding for 
infrastructure improvements that would benefit intercity rail 
passenger service. By contrast, the NGS experience shows that 
restructuring passenger train operations on freight railroad-owned 
lines without access to significant federal capital funding has grown 
increasingly difficult, given the capacity constraints on some lines 
and the limited availability of alternative sources of funding. 

That reality, which is not addressed in the report, explains many of 
the hurdles Amtrak encountered in implementing the NGS proposals. 

II. Capital Requirements: 

The report criticizes Amtrak for failing to quantify the capital 
investments required for each NGS-proposed route or service change, 
and for failing to weigh, in deciding what route and service changes 
to propose, the likelihood that particular states would provide 
capital funding. Both criticisms are unfounded. 

With regard to quantifying capital needs, the infrastructure 
investment required to add one or two train frequencies to a rail line 
varies considerably and is not easily quantifiable. In many cases, 
Amtrak has been able to initiate significant service changes without 
any capital investment, often by making minor modifications in its 
plans to address the initial objections that freight railroads almost 
invariably raise. Recent examples include the "Pennsylvanian" 
extension and the additional "Texas Eagle" frequencies discussed 
above. [See comment 7] 

Where infrastructure investment would likely be needed, the 
development of reliable cost estimates before Amtrak communicated its 
proposals to Congress would have required expensive and time consuming 
studies of the physical characteristics of each line, and of the 
number, type and schedules of the trains that operate over it. This 
effort would, of course, have significantly delayed issuance of the 
NGS Report. 

Moreover, as Amtrak explained to GAO — in a statement that appears, 
completely out-of-context, in the report — estimation of capital 
expenditures would have served no purpose even if Amtrak had unlimited 
time and resources. The purpose of the NGS Project was to identify all 
changes in Amtrak's routes and services that would more than cover 
their operating costs, and thus reduce Amtrak's operating losses and 
contribute to the statutory objective of self-sufficiency. If a 
potential route or service change passed muster under this test, it 
was in Amtrak's interest to propose it, even if it appeared likely to 
require capital investment, and to then pursue various means of 
funding that investment. [See comment 8] 

* In some cases, Amtrak expected that states and localities would fund 
capital expenditures associated with NGS proposals. As discussed 
below, this expectation was realized with regard to two NGS proposals. 

* In other cases, Amtrak reasonably anticipated that the expansion of 
its express-carrying capacity that would result from implementation of 
the NGS would create opportunities for mutually beneficial 
"gainsharing" arrangements, under which Amtrak would share express 
revenues in return for agreements that Amtrak could implement NGS 
proposals. Months of negotiations with two of the four largest freight 
railroads, which involved six meetings with one railroad and nearly a 
dozen meetings with the other, demonstrate that there was significant 
freight railroad interest in this proposal. However, the negotiations 
ultimately stalemated, though not on the issue of implementation, but 
rather on price. [See comment 9] 

* Finally, Amtrak could provide capital contributions in amounts 
negotiated with the host freight railroad, either by investing 
federally appropriated capital funds or through innovative funding 
mechanisms (such as the RRTF loan approach that Amtrak developed in 
conjunction with Kansas City Southern). Only under this scenario would 
it be necessary for Amtrak to negotiate a level of capital investment 
with the host freight railroad. However, in those negotiations, the 
preliminary estimates of required capital expenditures that GAO 
suggests Amtrak should have developed would, inevitably, have served 
as a floor on what the freight railroad would expect Amtrak to pay, 
even if better information suggested that those estimates were 
overstated. Clearly, from a bargaining perspective, that would not 
have been in Amtrak's interest. [See comment 10] 

With regard to states, the report claims that the restructuring of 
Amtrak's services in Florida was delayed by Amtrak's failure to 
ascertain, before issuance of the NGS Report, what level of capital 
funding Florida might provide. To the contrary, the state funding 
required for the Florida restructuring would never have materialized 
if Amtrak had followed GAO's suggested approach. [See comment 11] 

When Amtrak proposed the very ambitious Florida restructuring in the 
NGS Report, it appeared unlikely that the state would provide the 
significant capital funding required. However, Amtrak's announcement 
of its proposal created a groundswell of support among local 
communities that would gain Amtrak service, and their efforts to 
convince Florida's Governor and Department of Transportation to change 
their positions and support the necessary state funding were recently 
successful.[Footnote 5] 

Similarly, the NGS proposal to extend the then recently-instituted 
"Kentucky Cardinal" to Louisville, KY prompted the City of Louisville 
to allocate $300,000 in capital funds for this purpose, which Amtrak 
agreed to match. Service to downtown Louisville commenced in December 
of last year, and discussions are underway about a further extension 
of the service to Nashville. Here again, the necessary funding 
materialized only after — and indeed, only because — Amtrak publicized 
its proposal. 

III. Express Revenue and Cost Projections: 

The GAO is correct in stating that Amtrak overestimated the revenues 
and underestimated the costs attributable to handling additional 
express traffic on NGS-proposed trains. As discussed above, at the 
time the NGS Project was initiated, Amtrak's express initiative was in 
its infancy, and very little data on demand and costs was available. 
However, as the GAO notes, once it became apparent that net revenues
from express had been overestimated, Amtrak modified its projections, 
which led to a decision not to initiate some of the NGS-proposed 
services. 

The report acknowledges that, since the NGS Report was issued, Amtrak 
has made major changes in the management of its express operation, and 
in how it projects demand for and tracks costs of its express 
services. More specifically: 

(i) Amtrak has hired new senior managers in the mail and express group 
(including the current head of that group) who have substantial 
experience in premium cargo transportation, and a financial 
professional with 15 years experience to serve as the group's Chief 
Financial Officer; 

(ii) Amtrak has placed mail and express in a separate business unit 
with its own financial statements (developed on the same basis as the 
financial statements of Amtrak's other business units) and a new 
accounting structure to accurately track costs; 

(iii) Amtrak has adopted more reliable means for projecting express 
revenues and costs that are based upon the experience that Amtrak has 
gained in this business since the MBNA Project was initiated more than 
three years ago; and; 

(iv) Amtrak has also completed the time and motion studies at major 
terminals that are noted in the report. 

Amtrak believes that these actions address the issues regarding 
express revenues and costs that the report raises. 

Conclusion: 

Amtrak believes that the approach it followed in developing the NGS 
route and service change proposals was both reasonable and necessary. 
The NGS Report was not merely an action plan based upon rigorous 
financial analysis. It was also a vision of how Amtrak's national 
network could be reshaped so as to extend its reach and reduce 
operating losses, but without significant federal capital investment 
and with a statutory deadline for "operational self-sufficiency" 
looming just ahead. While the GAO criticizes Amtrak for failing to 
achieve all of its vision, we believe that Amtrak should not be 
faulted for pursuing an innovative approach to reconciling these 
conflicting objectives. [See comment 12] 

Sincerely, 

Signed by: 

Joe McHugh: 
Acting Vice President: 
Government Affairs: 

Footnotes: 

[1] "Report to Congress: The Market Based Network Analysis of the 
National Railroad Passenger Corporation", February 28, 2000 ("NGS 
Report"). 

[2] In light of the Inspector General's findings regarding RPS, and 
the fact that RPS was not utilized to develop the NGS proposals, 
Amtrak is puzzled by the inclusion in the report of a table that shows 
"Financial Performance of Amtrak's Routes" based upon RPS. [See 
comment 2] 

[3] The MBNA models were extensively analyzed by DOT's Inspector 
General, who received expert assistance from the Volpe National 
Transportation Center. In his "2000 Assessment of Amtrak's Financial 
Performance and Requirements", Report No. CR-2000-121, DOT's Inspector 
General concluded (at p. 30) that "[t]he MBNA models will generate 
more accurate projections of the revenue and expense impacts from fare 
and service changes than have previously been available to support 
Amtrak's decision-making." 

[4] It also bears noting that the Rail Passenger Service Act 
contemplates that issues regarding Amtrak route and service changes 
will be resolved quickly. The Act's provision (49 U.S.C. 24308(a)) 
that gives Amtrak a legal remedy before the Surface Transportation 
Board ("STB") should a freight railroad reject Amtrak's request to 
initiate a new route requires the STB to render its decision as to 
whether Amtrak should receive access within 90 days. 

[5] Contrary to the assertion in the GAO report, there is no 
inconsistency between Amtrak's and Florida's expectations regarding 
Amtrak's contribution to the Florida restructuring. The $61 million in 
state funds referenced in the report will cover virtually all 
improvements to the affected rail lines (with FEC funding the 
remainder), as well as the state's share of the costs of constructing 
stations (which will be matched by the local communities). Amtrak is 
not contributing money to third parties because it will be paying 
directly the capital costs it will cover (for improvements at Amtrak-
owned facilities and to Amtrak-owned cars and locomotives, and for 
certain start-up costs). 

The following is our evaluation of aspects of Amtrak’s comments on our 
draft report. 

1. We agree with Amtrak’s statement that a Senate committee report 
GAO’s Evaluation directed it to report on its conclusions regarding 
route and service changes before Amtrak issued its fiscal year 2000 
strategic business plan. However, nothing in the Senate report 
required Amtrak to issue its strategic business plan by a certain date 
or earlier than it would have done otherwise. As such, the Senate 
report language did not create the sense of urgency that Amtrak 
implies. 

2. Amtrak questioned our including a list of route profits and losses 
in appendix I of the draft report because the route profitability 
system used to generate these results does not produce accurate 
information for making route decisions. We are not suggesting that the 
route profitability statistics should have been used in making Network 
Growth Strategy decisions. We are also not suggesting that they should 
be used, by themselves, in making future route decisions. Other 
metrics should also be employed.[Footnote 18] However, the fact that 
Amtrak loses money on nearly every single route that it operates (for 
example, 20 of Amtrak’s 43 routes lost more than $20 million in 2001, 
even after including state support) was the basis for Amtrak deciding 
to contract routes in 1995 and expand them in 1998. As such, this 
route profitability information is contextually important in Amtrak’s 
quest to improve its financial condition. 

3. Amtrak interpreted our conclusion on the need for early 
consultation with freight railroads before it announced its Network 
Growth Strategy to mean that we were advocating that it should have 
“engaged in lengthy ‘consultations’ with each of the affected 16 
freight railroads” and conducted “expensive and time consuming studies 
of the physical characteristics of each line, and of the number, type, 
and schedules of the trains that operate over it.” We did not intend 
such an interpretation. We agree with Amtrak that some proposed route 
and service changes might be expected to be easier to implement than 
others—such as ones that could be expected to have little effect on 
freight railroads. We also agree with Amtrak that there is no model 
for how and on what timetable such issues should be resolved. We do 
not prescribe a level of specificity for these discussions, specific 
expected outcomes, or uniformity. We believe that discussions with 
freight railroads should be tailored to the complexity, expected 
difficulty, and risk associated with each proposed action. As 
discussed previously, we revised our recommendation to make it more 
useful to Amtrak. 

4. Amtrak stated that “in many cases [it] was able to implement 
significant route and service changes fairly quickly” and cited two 
examples. In reality, Amtrak has implemented only 3 of the 15 planned 
route and service actions. 

5. Amtrak states that our draft report implied that it had no reason 
to expect Norfolk Southern would agree to the operation of the Skyline 
service. Amtrak disagreed with our draft report because it had 
implemented a similar service (an extension of the Pennsylvanian) over 
the same route about a year before the Network Growth Strategy was 
issued. Any implication about the potential success or failure in 
implementing the Skyline service was inadvertent. We have revised our 
report to state that Norfolk Southern was willing to work with Amtrak 
to establish this service and had reached agreement on how the train 
was to be handled. The report states that Amtrak largely canceled this 
proposal because it and Norfolk Southern could not agree on 
compensation for Amtrak’s use of Norfolk Southern’s tracks. 

6. Amtrak disagreed with the example we used to illustrate what occurs 
when early discussions with freight railroads do not occur. Amtrak 
stated that its Network Growth Strategy contemplated using either 
Union Pacific or Kansas City Southern tracks and Amtrak approached 
both railroads and that its decision to re-route the Crescent Star 
from Union Pacific Railroad to Kansas City Southern tracks represents 
a prudent business decision rather than a flaw in its decisionmaking. 
We agree that the Network Growth Strategy provided flexibility in 
routing and that deciding to re-route the Crescent Star might have 
been a prudent business decision. However, Amtrak did not learn until 
spring 2000 that (1) significant capital improvements were required to 
implement the service, (2) Union Pacific was not willing to share the 
capital investment costs needed to use this line, and (3) an 
alternative routing would be required. Since Amtrak planned to 
implement the Crescent Star in summer 2000, just a few months after it 
announced the Network Growth Strategy, having early knowledge of 
significant potential roadblocks would have been useful to Amtrak—-for 
example, either in attempting to ameliorate the roadblocks or deciding 
earlier to concentrate on the alternative Kansas City Southern route. 
As Amtrak stated in its comments, it needed to implement the Network 
Growth Strategy quickly to help reduce its need for operating 
subsidies. 

7. Amtrak commented that (1) the infrastructure investment required to 
add one or two train frequencies to a rail line is not easily 
quantifiable and (2) there are other ways to reach agreements to 
undertake capital projects other than by allocating costs between 
parties (e.g., the freight railroad might agree to bear the cost of 
the project if Amtrak agrees to something else). We agree with Amtrak’
s statement. However, a recurring theme for Amtrak has been its dearth 
of capital to improve its service. We believe that it would have been 
prudent for Amtrak to factor into its decisionmaking the fact that 
capital issues, for some proposed routes, were crucial to Amtrak’s 
being able to implement the Network Growth Strategy, particularly as 
it recognized in its comments that “…freight railroads can, not 
infrequently, demand large infusions of capital from passenger train 
operators to accommodate additional trains.” 

8. Amtrak states that our discussion of capital funding is out of 
context because the purpose of the Network Growth Strategy was to 
identify routes for which revenues would exceed operating costs. 
Amtrak stated that if a potential route or service change met this 
test then it made sense to pursue it, even if it was likely to require 
capital support. We agree with Amtrak that it made sense to pursue 
routes that were contemplated to make an operating profit even if 
capital investment would be needed to implement them. We did not 
intend to suggest that Amtrak should have pursued only route and 
service expansions that were likely to cover both operating and 
capital costs. Rather, we believe that, for some routes, capital 
investment was an important prerequisite to Amtrak being able to 
implement the routes quickly so that it could obtain the operating 
profits. 

9. Amtrak commented that some Network Growth Strategy negotiations 
with freight railroads were stalemated not on the issue of 
implementation, but on price. We have revised our report to recognize 
this. 

10. Amtrak criticized our suggestion that it should have had 
preliminary discussions with freight railroads over capital funding 
issues, saying that it is a poor negotiating technique to approach a 
freight railroad by telling it how much capital Amtrak is willing to 
contribute, because this figure sets a floor for Amtrak’s 
contribution. We agree that there are delicate business issues arising 
from Amtrak’s attempts to get freight railroads to allow it to expand 
operations over freight railroad-owned tracks and that different route 
and service proposals can raise different and sometimes complex 
issues. We are not suggesting that Amtrak “give away the store” in 
these discussions by disclosing in advance how much capital support it 
might be willing to contribute to the freight railroads. However, as 
discussed in the report, we believe that it would have been prudent to 
determine freight railroads’ expectations before deciding to implement 
the plan because freight railroads’ cooperation was imperative to the 
success of the Network Growth Strategy. Without an understanding of 
whether freight railroads’ expectations were similar to Amtrak’s—and 
the expected ease or difficulty in meshing these expectations—Amtrak 
had little basis to expect that the route and service proposals it 
made could actually be implemented expeditiously so that they could 
help reduce Amtrak’s operating losses. 

11. On several bases Amtrak disagreed with our discussion of gaining 
an early understanding of whether states, such as Florida, might or 
might not be willing to provide the capital funds that Amtrak expected 
them to contribute. Because the focus of our work was Amtrak’s 
interaction with freight railroads, we have deleted references to the 
capital support that Amtrak expected from states such as Florida. 

12. Amtrak stated that the Network Growth Strategy was not just “…an 
action plan based on rigorous financial analysis. It was a vision of 
how Amtrak’s national network could be reshaped so as to extend its 
reach and reduce operating losses…” [emphasis in the original]. Amtrak 
suggested that we faulted it for pursuing an innovative approach and 
because it did not achieve “all its vision.” We are not criticizing 
Amtrak for pursuing a route expansion strategy. Rather, our report 
focuses on the aspects that might have made the vision more successful 
than it was, although perhaps at a more modest level than Amtrak 
originally envisioned. Amtrak’s Network Growth Strategy ultimately 
failed because the route system expanded marginally and Amtrak was not 
able to reduce its operating losses to the extent planned. In our 
opinion, an important contributor to this failure was Amtrak’s 
inattention to potential implementation problems before it announced a 
strategy. Attention to potential implementation problems was crucial 
because, as Amtrak stated, it needed to have the routes implemented 
quickly so as to reap the financial benefits that would result in a 
reduction of operating losses. We believe that the recommendation we 
offer, if adequately implemented, could help Amtrak be more successful 
in any future route expansion efforts. 

[End of section] 

Footnotes: 

1. In December 1994, at the direction of the administration, Amtrak 
established the goal of eliminating its need for federal operating 
assistance by December 2002 (called “operational self-sufficiency”). 
This goal became a requirement under the Amtrak Reform and 
Accountability Act of 1997. 

2. Route actions can extend, truncate, launch, or close routes. 
Service changes generally increase or decrease the frequency with 
which trains run on a route. 

3. Express is the transportation of higher-value, time-sensitive 
merchandise, such as food and automobile parts. This plan also 
included actions to improve efficiencies, such as redeploying cars and 
locomotives among routes. 

4. This amount is measured on a cash basis—that is, revenues less cash 
expenses. Depreciation is not included. Another route, the Heartland 
Flyer between Texas and Oklahoma, made a profit of $600,000 primarily 
because the state of Oklahoma provided Amtrak with $4.6 million, or 
about 80 percent of the route’s total revenue. 

5. It also launched four new routes and reinstated one route. See 
appendix II for a list of route actions for the period from 1995 
through 2001. 

6. U.S. General Accounting Office, Intercity Passenger Rail: Financial 
Performance of Amtrak’s Routes, GAO/RCED-98-151 (Washington, D.C.: May 
14, 1998). 

7. See appendix IV for a description of how Amtrak evaluates route and 
service proposals and appendix V for a description of the models 
Amtrak uses to estimate route-related ridership, revenues, and costs. 

8. Typically more than one reason led to the decision to cancel a 
route. For example, some routes were canceled in part because Amtrak 
was not able to secure the mail and express business it planned and in 
part because it could not reach agreement with freight railroads over 
who would pay for capital improvements or about compensation for 
Amtrak operating its trains over freight railroad tracks. 

9. The president of Amtrak’s mail and express unit told us that Amtrak 
has recently begun time and motion studies at certain terminals to 
better understand the true costs of mail and express activities. 

10. In more limited circumstances Amtrak approached states about 
contributing capital funds. 

11. An official from Norfolk Southern Corporation said that his 
railroad did have detailed discussions with Amtrak in early 2000. 
However, he said that the only discussion about capital investments 
involved low-cost items. 

12. Amtrak does consider capital needs when high-speed rail corridors 
are involved. See appendix V. 

13. Materials supplied to Amtrak’s board of directors for its use in 
considering the adoption of the Network Growth Strategy indicated that 
capital improvements would likely be needed for portions of six of the 
proposed routes, including the Crescent Star. However, the material 
did not indicate the amount of capital that would be needed or how 
significant availability of capital was to the implementation of the 
proposals. 

14. According to Amtrak officials, the Crescent Star service has not 
been implemented because of continuing negotiations over the 
collateral requirements of a federal loan. 

15. A Union Pacific official also commented that Amtrak could have 
encountered Customs Service issues on this route. 

16. The strategic business units are Intercity, Amtrak West, and 
Northeast Corridor. The Mail and Express unit became an independent 
business unit in 2001. 

17. Ten high-speed rail corridors (with train speeds of at least 90 
miles per hour) have been designated either by legislation or by the 
Department of Transportation. These corridors are in various stages of 
planning. An eleventh corridor is Amtrak’s Northeast Corridor. 

18. For a discussion of initial considerations that could guide 
Congress as it debates the future role of the federal government in 
supporting intercity passenger rail, see U.S. General Accounting 
Office, Intercity Passenger Rail: Congress Faces Critical Decisions in 
Developing a National Policy, GAO-02-522T (Washington, D.C.: Apr. 11, 
2002). 

[End of section] 

GAO’s Mission: 

The General Accounting Office, the investigative arm of Congress, 
exists to support Congress in meeting its constitutional 
responsibilities and to help improve the performance and accountability 
of the federal government for the American people. GAO examines the use 
of public funds; evaluates federal programs and policies; and provides 
analyses, recommendations, and other assistance to help Congress make 
informed oversight, policy, and funding decisions. GAO’s commitment to 
good government is reflected in its core values of accountability, 
integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through the Internet. GAO’s Web site [hyperlink, 
http://www.gao.gov] contains abstracts and fulltext files of current 
reports and testimony and an expanding archive of older products. The 
Web site features a search engine to help you locate documents using 
key words and phrases. You can print these documents in their entirety, 
including charts and other graphics. 

Each day, GAO issues a list of newly released reports, testimony, and 
correspondence. GAO posts this list, known as “Today’s Reports,” on its 
Web site daily. The list contains links to the full-text document 
files. To have GAO e-mail this list to you every afternoon, go to 
[hyperlink, http://www.gao.gov] and select “Subscribe to daily E-mail 
alert for newly released products” under the GAO Reports heading. 

Order by Mail or Phone: 

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to: 

U.S. General Accounting Office: 
441 G Street NW, Room LM: 
Washington, D.C. 20548: 

To order by Phone: 
Voice: (202) 512-6000: 
TDD: (202) 512-2537: 
Fax: (202) 512-6061: 

To Report Fraud, Waste, and Abuse in Federal Programs Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Public Affairs: 

Jeff Nelligan, managing director, NelliganJ@gao.gov: 
(202) 512-4800: 
U.S. General Accounting Office: 
441 G Street NW, Room 7149:
Washington, D.C. 20548: 

[End of document]