This is the accessible text file for GAO report number GAO-13-304 
entitled 'California High-Speed Passenger Rail: Project Estimates 
Could Be Improved to Better Inform Future Decisions' which was 
released on March 29, 2013. 

This text file was formatted by the U.S. Government Accountability 
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 Government Accountability Office: 
GAO: 

Report to Congressional Requesters: 

March 2013: 

California High-Speed Passenger Rail: 

Project Estimates Could Be Improved to Better Inform Future Decisions: 

GAO-13-304: 

GAO Highlights: 

Highlights of GAO-13-304, a report to congressional requesters. 

Why GAO Did This Study: 

The planned 520-mile California high-speed rail project, which would 
link San Francisco to Los Angeles, would be designed to operate at 
speeds up to 220 miles per hour. At an estimated cost of $68.4 billion 
(in year-of-expenditure dollars), it is expected to be one of the most 
expensive transportation projects undertaken in the United States. The 
Authority is responsible for implementing the project and federal 
funding is being provided from the FRA’s High-Speed Intercity 
Passenger Rail program. GAO reviewed (1) the reliability of project 
cost estimates, (2) the reasonableness of revenue and passenger rail 
ridership forecasts, (3) the risks attendant with the project’s 
funding plan, and (4) the comprehensiveness with which the project’s 
economic impacts were identified. 

GAO obtained documents from and conducted interviews with federal 
officials and officials from the Authority related to cost, financing, 
ridership and revenue modeling and estimation, and business plans and 
analyses related to potential economic impacts. GAO also interviewed 
state and local officials as well as the project’s peer review group 
members. 

What GAO Found: 

The California High-Speed Rail Authority (Authority) met some, but not 
all of the best practices in GAO’s Cost Estimating and Assessment 
Guide (Cost Guide) for producing cost estimates that are accurate, 
comprehensive, well documented, and credible. By not following all 
best practices, there is increased risk of such things as cost 
overruns, missed deadlines, and unmet performance targets. The 
Authority substantially met the criteria for the accurate 
characteristic by, for example, the cost estimate’s reflecting the 
current scope of the project. However, the Authority partially met the 
criteria for the other three characteristics since the operating costs 
were not sufficiently detailed (comprehensive), the development of 
some cost elements were not sufficiently explained (well documented), 
and because no systematic assessment of risk was performed (credible). 
The Federal Railroad Administration (FRA) issued limited guidance for 
preparing cost estimates, and this guidance did not reflect best 
practices in the Cost Guide. The Authority plans to improve its cost 
estimates. 

GAO found the Authority’s ridership and revenue forecasts to be 
reasonable; however, additional updates are necessary to refine the 
ridership and revenue model for the 2014 business plan. GAO also found 
the travel-demand-modeling process used to generate these forecasts 
followed generally accepted travel-demand-modeling practices. For 
example, the Authority revised several assumptions, such as gasoline 
price forecasts, to reflect changes in current and anticipated future 
conditions. However, additional updates, such as the development of a 
new travel survey, will be necessary to further refine these forecasts 
and improve the model’s utility to make future decisions. External 
peer review groups have also recommended additional updates. 

The project’s funding, which relies on both public and private 
sources, faces uncertainty, especially in a tight federal and state 
budget environment. Obtaining $38.7 billion in federal funding over 
the construction period is one of the biggest challenges to completing 
this project. In the latter stages, the Authority will also rely on 
$13.1 billion in private-sector financing, but will require more 
reliable operating cost estimates and revenue forecasts to determine 
whether, or the extent to which, the system will be profitable. The 
Authority’s plan recognizes the uncertainty of the current funding 
environment and is building the project in phases. The Authority has 
also identified an alternative funding source. However, that funding 
source is also uncertain. 

The Authority did a comprehensive job in identifying the potential 
economic impacts of the high-speed rail project. This includes 
identification of user impacts, such as effects on travel time 
reliability, and non-user impacts, such as effects on highway 
congestion. However, the nature of specific economic impacts will 
depend on a number of factors, including future project decisions. GAO 
also found limitations in the Authority’s benefit-cost analysis of the 
project that could limit its usefulness to decision makers. Finally, 
GAO found that construction of the high-speed rail project will not 
eliminate the need for additional improvements to meet future 
statewide-travel demand, but current statewide-transportation 
assessments and planning have given little consideration to this issue. 

What GAO Recommends: 

To produce reliable cost estimates, FRA should improve its guidance so 
it is in line with the best practices in GAO’s Cost Guide. The 
Department of Transportation did not agree or disagree with the 
recommendation but said, with further analysis, applying the Cost 
Guide would be feasible. The Authority said it will incorporate many 
of the report’s findings into future cost and ridership estimates. 

View [hyperlink, http://www.gao.gov/products/GAO-13-304]. For more 
information, contact Susan Fleming at (202) 512-2834 or 
flemings@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

Authority Substantially or Partially Met GAO's Best Practices for 
Producing Reliable Cost Estimates, but Can Make Improvements: 

Ridership and Revenue Forecasts Are Reasonable for Current Purposes, 
but Will Require Updates: 

The Authority's Funding Plan Faces Uncertainty: 

The Authority Comprehensively Identified Potential Economic Impacts, 
but It Is Too Early to Determine Specific Impacts: 

Conclusion: 

Recommendation: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: GAO Analysis of California High-Speed Rail Authority's 
Cost Estimates: 

Appendix III: Description of Generally Accepted Travel-Demand-Modeling 
Practices and Authority's Methods for Developing Ridership and Revenue 
Forecasts: 

Appendix IV: Comments from the California High-Speed Rail Authority: 

Appendix V: GAO Contact and Staff Acknowledgment: 

Tables: 

Table 1: Summary of GAO's Assessment of California High-Speed Rail 
Authority's Cost Estimates: 

Table 2: Description of Updated California High-Speed Rail Travel-
Demand-Model Assumptions: 

Table 3: Funding Secured for Constructing the California High-Speed 
Rail Project, as of February 2013: 

Table 4: The Authority's Funding Plan for Constructing Phase 1 of the 
California High-Speed Rail Project, as of April 2012: 

Table 5: Primary Economic Impacts Associated with High-Speed Rail 
Identified by the Authority in the April 2012 Revised Business Plan, 
BCA, or EIA: 

Table 6: Organizations and Individuals Interviewed: 

Table 7: Guidance and Reports Used to Identify Generally Accepted 
Travel-Demand-Modeling Practices: 

Figures: 

Figure 1: Map of Planned California High-Speed Rail System and Key 
Construction Phases, 2013: 

Figure 2: Time Line of California High-Speed Rail Project: 

Abbreviations: 

AAPOR: American Association for Public Opinion Research: 

ASC: Aviation System Consulting, LLC: 

BCA: benefit-cost analysis: 

CTC: California Transportation Commission: 

DOT OIG: Department of Transportation Office of Inspector General: 

EIA: economic impact analysis: 

FHWA: Federal Highway Administration: 

FRA: Federal Railroad Administration: 

FTA: Federal Transit Administration: 

HSIPR: High Speed Intercity Passenger Rail: 

ICE: independent cost estimate: 

IOS: initial operating segment: 

ITS: Institute of Transportation Studies: 

LAO: Legislative Analyst's Office: 

LOS: level of service: 

MPO: Metropolitan Planning Organization: 

MTC: Metropolitan Transportation Commission: 

NEC: Northeast Corridor: 

NEPA: National Environmental Policy Act: 

NOFA: Notice of Funding Availability: 

O&M: operations and maintenance: 

OMB: Office of Management and Budget: 

PRG: peer review group: 

PRIIA: Passenger Rail Investment and Improvement Act of 2008: 

SCC: Standard Cost Categories: 

TIGER: Transportation Investments Generating Economic Recovery: 

UIC: International Union of Railways: 

WBS: work breakdown structure: 

YOE: year of expenditure: 

[End of section] 

United States Government Accountability Office: 
Washington, DC 20548: 

March 28, 2013: 

Congressional Requesters: 

California's high-speed rail project is poised to be the first rail 
line in the United States designed to operate at speeds up to 220 
miles per hour.[Footnote 1] At an estimated cost of $68.4 billion (in 
year-of-expenditure dollars), it is expected to be one of most 
expensive transportation projects undertaken in the United States. 
[Footnote 2] The Federal Railroad Administration (FRA) has obligated 
nearly $3.5 billion in grants to begin construction later this year. 
To complete the project, the state of California expects to rely on 
significant public-sector funding, in addition to private funding, 
through the project's anticipated completion date in 2028. 

Over the past two decades, there has been interest in developing high-
speed rail between San Francisco and Los Angeles. Rail proponents 
argue that California faces significant transportation congestion and 
that high-speed rail can help alleviate the need to expand highway and 
air infrastructure. Proponents also argue that the corridor has 
characteristics that make high-speed rail viable: it connects highly 
populated metropolitan centers along a corridor that is currently 
highly traveled by air, rail, and automobile passengers. Opponents of 
the project argue that it is too expensive, especially with tight 
federal and state budgets, and has not sufficiently demonstrated that 
the project will be financially viable. As we reported in 2009, 
completing a high-speed rail project requires significant and 
sustained political, public, and financial support given the high 
costs and multiple stakeholders involved.[Footnote 3] In recent years, 
legislation has been enacted at the state and federal level that has 
allowed California's project to move forward by appropriating funding 
for high-speed rail. In 2008, California's voters approved a $9.95-
billion bond measure, and in 2012, the state legislature appropriated 
over $4 billion from this bond to the project and connecting transit 
upgrades. At the federal level, the Passenger Rail Investment and 
Improvement Act of 2008 (PRIIA) authorized development of high-speed 
intercity passenger-rail corridors and the American Recovery and 
Reinvestment Act of 2009 (Recovery Act) appropriated $8 billion to 
fund development of these corridors and intercity passenger-rail 
projects.[Footnote 4] In June 2009, the FRA established the High Speed 
Intercity Passenger Rail (HSIPR) program that provides discretionary 
grants for high-speed or intercity passenger rail projects. In 
December 2009, the Fiscal Year 2010 Department of Transportation (DOT) 
appropriations act appropriated $2.5 billion for the HSIPR program 
($400 million of which was later rescinded).[Footnote 5] 

The California High-Speed Rail Authority (Authority)--a state agency 
formed in 1996--has been tasked with implementing the project and has 
published four successive business plans detailing the project's cost 
estimates, financing plan, ridership projections, and anticipated 
economic impacts. Since 2009, the scope of the project has changed 
along with the project's estimated costs. In addition, questions have 
been raised about how the project will be funded beyond the current 
$11.5 billion in state and federal funding commitments, as well as the 
project's ability to attract sufficient ridership and bring economic 
benefits to the cities and communities along the corridor. This report 
assesses: (1) the reliability of the Authority's estimates of the 
project's costs, (2) the reasonableness of the Authority's passenger 
rail ridership and revenue forecasts, (3) the risks attendant with the 
Authority's funding plan for the project, and (4) the 
comprehensiveness with which the Authority identified potential 
economic impacts of the project. As such, we are assessing the quality 
of the available information used by policymakers. We are not 
evaluating the merits of the project itself, which should be 
considered in light of whether this project best meets the 
transportation needs of the estimated 51 million Californians in 2050. 

We obtained information from numerous sources to address our 
objectives, including the models underlying the various estimates and 
interviews with Authority officials and contractors that prepared 
them, the peer review group members that reviewed them, outside 
critics and supporters, high-speed rail experts and FRA officials. We 
based our analysis on estimates underlying the Authority's April 2012 
revised business plan. 

To assess the reliability of the project's cost estimates, we compared 
the Authority's cost estimating approach to GAO's best practices found 
in the 2009 GAO Cost Estimating and Assessment Guide (Cost Guide). 
[Footnote 6] We used the Cost Guide to conduct an in-depth evaluation 
of the methods and assumptions used by the Authority to develop cost 
estimates of Phase 1 of the project; our goal was to determine whether 
best practices were employed that would help ensure that the cost 
estimates are well-documented, comprehensive, accurate, and credible. 
The Cost Guide was developed for cost estimators to assist them in 
preparing reliable estimates of capital program costs. While GAO's 
Cost Guide provides criteria to evaluate the Authority's methods for 
developing its cost estimates, we cannot use it to determine whether 
or not the Authority's cost estimate is precise. Nor can we use the 
Cost Guide to evaluate the benefits of this project in comparison to 
other transportation improvements. Such an evaluation would require a 
cost-benefit and other analyses, which we did not conduct as part of 
this work. To assess the reasonableness of the ridership and revenue 
estimates, we analyzed the extent to which the Authority's ridership 
model methodology adhered to FRA guidance and generally accepted 
practices and reviewed peer review reports assessing the model's 
methodology. We also interviewed Authority officials, their 
contractor, and peer review panel members. To assess the Authority's 
financing plan, we reviewed the plan, conducted interviews with 
Authority and other state and federal officials, and reviewed 
literature and other information on financing for high-speed rail 
projects in other countries as well as large transportation projects 
in the United States. To assess the comprehensiveness with which the 
Authority identified potential economic impacts of the project, we 
compared the impacts identified in the April 2012 revised business 
plan and related economic analyses prepared by the Authority with 
criteria for such assessments contained in DOT's program-funding 
notices, applicable legislation, and best practices found in academic 
literature and federal guidelines. For purposes of this report, 
potential economic impacts include those impacts on users of a 
transportation system such as travel time savings and those impacts to 
non-users of a system such as highway congestion or noise levels. (For 
more detailed information on our objectives, scope, and methodology, 
see appendix I.) 

We conducted this performance audit from February 2012 to March 2013 
in accordance with generally accepted government auditing standards. 
These standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

Background: 

In 2008, California voters approved Proposition 1A,[Footnote 7] which 
authorized $9.95 billion in state bond funding for construction of the 
California high-speed rail system and connection improvements to 
existing passenger rail systems. Proposition 1A established several 
requirements for this high-speed rail system, such as that the rail 
system must be capable of sustained operating speeds of no less than 
200 miles per hour, and once built, must operate without a public 
subsidy. The planned 520-mile high-speed rail system will operate 
between San Francisco and Los Angeles at speeds up to 220 miles per 
hour (see figure 1). The Authority is the state entity charged with 
planning, designing, and constructing the California high-speed rail 
system. The Authority has a nine-member policy board appointed by the 
California legislature and Governor, and a staff of approximately 55 
state employees who oversee, among other things, contracts for 
environmental review, preliminary engineering design, preliminary 
right-of-way acquisition tasks,[Footnote 8] contractor oversight and 
other activities.[Footnote 9] 

Figure 1: Map of Planned California High-Speed Rail System and Key 
Construction Phases, 2013: 

[Refer to PDF for image: illustrated map] 

First construction: 
North of Fresno to just north of Bakersfield: 130 miles; 
Estimated completion date: 2018. 

Initial operating segment (IOS): 
Merced to San Fernando Valley: 300 miles; 
Estimated completion date: 2021. 

Bay to Basin: 
San Jose to San Fernando Valley: 410 miles; 
Estimated completion date: 2026. 

Phase 1 blended: 
* San Francisco to Los Angeles: 520 miles; 
* Integrated with Caltrain system in San Jose; 
* Dedicated high-speed rail tracks to Los Angeles Union Station; 
* Multimodal connection with Metrolink at Palmdale; 
Estimated completion date: 2028. 

Sources: California High Speed Rail Authority and GAO. 

[End of figure] 

Construction of the California high-speed rail project is expected to 
occur in phases beginning with a 130-mile section from just north of 
Fresno, California, to just north of Bakersfield, California. 
Construction will begin in the Central Valley and proceed to other 
portions of the corridor as funding is available. The Central Valley 
is the furthest advanced in terms of design and engineering work, as 
well as environmental reviews. For example, FRA approved a preferred 
route alignment for the Merced to Fresno, California, portion of the 
corridor in September 2012. According to FRA, the federally funded 
portion of the project in the Central Valley has more complete cost 
estimates than subsequent segments given that preliminary engineering 
and environmental reviews are complete or nearly complete. Other 
project segments, however, are in different stages of development and 
have different levels of information from which to develop cost 
estimates. In July 2012, the California legislature appropriated $4.7 
billion of the $9.95 billion in state bond funds, including $2.6 
billion for construction of the high-speed rail project and $1.1 
billion for upgrades in the San Francisco peninsula and in the Los 
Angeles basin (commonly referred to as the "bookends").[Footnote 10] 
The process of acquiring property for the right-of-way and 
construction has begun. Requests for proposals to select construction 
contractors and right-of-way acquisition were issued in March and 
September 2012, respectively. In addition, in January 2013, the 
Authority awarded a project and construction management contract for 
the initial phase of California's high-speed rail project. According 
to the Authority, a design-build contract for the first construction 
(covering approximately 30 miles) is expected to be awarded in June 
2013 with construction planned to commence in summer 2013. (See figure 
2). 

Figure 2: Time Line of California High-Speed Rail Project: 

[Refer to PDF for image: time line] 

1996: 
The Authority was established by the state legislature to design and 
build the high-speed rail project. 

November 2008: 
California voters approved Proposition 1A, a bond referendum to 
provide $9.95 billion to fund initial design and construction of the 
project. 

2009: 
FRA established the HSIPR program, and Congress appropriated $8 
billion for high-speed rail projects as part of Recovery Act. 

2010: 
Congress appropriated $2.5 billion to the HSIPR program. 

November 2011: 
The Authority issued a draft business plan that envisions a full-build 
high-speed rail project at a total year-of-expenditure cost of $98.5 
billion. 

April 2012: 

Authority issued a revised business plan that envisions a blended high-
speed rail system and project cost is reduced to $68.4 billion in year-
of-expenditure dollars. 

July 2012: 
The California state legislature appropriated $4.7 billion to the 
project from Proposition 1A high-speed rail bonds. 

September 2017: 
In accordance with governing grant agreements, all Recovery Act funds 
must be expended. 

2022: 
IOS expected to be operational[A] from Merced to the San Fernando 
Valley. 

2026: 
Bay to Basin segment expected to be operational from San Jose to the 
San Fernando Valley. 

2029: 
Phase 1 blended system expected to be operational from San Francisco 
to Los Angeles. 

Source: GAO. 

[A] The estimated construction date refers to the date for which 
construction of a segment is expected to be completed while the 
expected date of operations refers to the date for which a segment is 
expected to be operational. The estimated construction dates for the 
IOS and Phase 1 are 2021 and 2028, respectively, which differs from 
the estimated operational dates which are 2022 and 2029, respectively. 

[End of figure] 

The federal government has committed funding to the project. The FRA 
awarded the state approximately $3.3 billion in capital construction 
funds and $231 million for environmental review and preliminary 
engineering work under the HSIPR program for a total of approximately 
$3.5 billion.[Footnote 11] The California high-speed rail project is 
the largest recipient of HSIPR funds, with about 35 percent of program 
funds obligated. Most of the HSIPR money awarded to the project was 
appropriated by the Recovery Act and, in accordance with governing 
grant agreements, must be expended by September 30, 2017.[Footnote 12] 
In addition, approximately $945 million in fiscal year 2010 funding 
was awarded to the project by FRA and is to remain available until 
expended. While the funds remain available until expended under FRA's 
fiscal year 2010 appropriation, the governing grant agreements specify 
the schedule for expenditure of funds. 

Even though some funding has been committed, additional funding will 
be needed to complete the project. For example, according to the 
Authority's finance plan, over $38 billion in federal funds and over 
$4 billion from Proposition 1A proceeds will be needed to complete 
Phase 1 of the project. In addition, the Authority is also planning to 
obtain another $13.1 billion in private-sector capital to help defray 
the cost of construction after the initial operating segment is 
completed in 2028. 

As the federal agency responsible for awarding and overseeing grants 
to HSIPR applicants, the FRA established guidance outlining 
requirements and procedures and developed an oversight program to 
ensure that the project's goals, objectives, performance requirements, 
budgets, and other related program criteria are being met. Thus far, 
FRA's guidance to HSIPR grant applicants has been limited with respect 
to developing cost estimates and ridership and revenue forecasts. The 
Department of Transportation's Office of Inspector General (DOT OIG) 
noted that the lack of clear, detailed guidance allows for analyses of 
widely varying quality, making it difficult to accurately assess 
whether projects will be viable or require substantial financial 
support and has recommended FRA improve its guidance.[Footnote 13] In 
addition, we have previously reported that a clear definition of the 
federal role, goals, and objectives in conjunction with a robust grant 
oversight program, is critical to FRA making sound federal investments 
in high-speed rail projects.[Footnote 14] 

The Authority is required to prepare and periodically submit to the 
state legislature a business plan, which must identify, among other 
things, ridership estimates, operating and maintenance costs, and the 
source of project funding.[Footnote 15] The Authority's next business 
plan is expected to be released in 2014. Several groups have been 
established to review and comment on the estimates presented in the 
Authority's business plans. For example, an independent peer review 
group (PRG) was established in accordance with California law to 
evaluate the Authority's funding plans and report its judgment as to 
the feasibility and the reasonableness of the plans, appropriateness 
of assumptions, analyses and estimates, and any observations or 
evaluations it deems necessary.[Footnote 16] In addition, the 
Authority convened a Ridership and Revenue Peer Review Panel (Panel) 
to review the Authority's ridership and revenue-forecasting process 
and outcomes and conduct an in-depth review of the models used to 
estimate ridership and revenue and the forecasts derived from them. 
The California state auditor is required to periodically audit the 
Authority's use of bond proceeds.[Footnote 17] 

In response to the initial high estimated cost of building the San 
Francisco to Los Angeles route--about $98 billion--and other 
criticisms of the Authority's November 2011 draft business plan, the 
project underwent substantial revision for the April 2012 revised 
business plan. Most significantly, the Authority scaled back its plans 
to build dedicated high-speed rail lines over the project's entire 
length. Instead, the April 2012 revised business plan adopted a 
"blended" system in which high-speed rail service would be provided 
over a mix of dedicated high-speed lines and existing and upgraded 
local rail infrastructure (entirely at the bookends of the system on 
the San Francisco peninsula and in the Los Angeles basin). The 
estimated cost in the April 2012 revised business plan is $68.4 
billion. 

The ridership and revenue forecasts in the April 2012 revised business 
plan also changed from the November 2011 plan. For example, in the 
November 2011 draft business plan, the Authority provided low and high 
estimates of ridership in 2030 of 14.4 million and 21.3 million 
passengers. In the April 2012 revised business plan, these estimates 
increased by 12 and 26 percent, respectively, to 16.1 million and 26.8 
million. Revenues similarly increased between the two plans from a low 
and high estimate of $1.05 billion and $1.56 billion in the November 
2011 plan[Footnote 18] to $1.06 billion and $1.81 billion in April 
2012.[Footnote 19] The ridership and revenue estimates increased 
because among other things, a "one-seat" service from San Francisco to 
Los Angeles would begin sooner under the blended approach than the 
original solely dedicated lines approach. However, by 2040 ridership 
forecasts under the blended approach are less than the original full 
build approach. The range between the high and low estimates also 
increased between the November and April plans reflecting a greater 
degree of uncertainty in the estimates. We have previously reported 
that forecasting ridership and revenue is a complex and iterative 
process and that early stage estimates should be based on the best 
available data and what is initially known about the proposed project. 
[Footnote 20] As additional information becomes available, the 
Authority's model used to produce the forecasts is intended to be 
updated. 

Development of the high-speed rail system has been controversial with 
many strongly held beliefs among the numerous supporters and opponents 
of the project. Supporters have cited the need for high-speed rail to 
address growing congestion concerns, particularly in the metropolitan 
areas, and to address future transportation demands. Supporters have 
noted that California's expected population growth--which is expected 
to increase from 38 million in 2012 to an estimated 51 million 
Californians in 2050--and economic growth will continue to place more 
demands on California's transportation infrastructure requiring that 
significant new capacity be added to its transportation network. 
Further, supporters argue that the cost of expanding the current 
network of highways and airports to meet current and future 
transportation needs is cost prohibitive and would be detrimental to 
air quality and that high-speed rail will increase economic 
development in local communities to be served by high-speed rail and 
generate new jobs. In addition, supporters also note that several 
critical airport and highway expansions are infeasible due to land 
constraints, particularly at key airports and urban segments of 
highways. Opponents of the plan have argued, among other things, that 
the cost of the high-speed rail system is too great and that future 
funding for the system is too uncertain given the current fiscal 
environment. Opponents have also raised concerns about the credibility 
of the ridership and revenue forecasts presented in the Authority's 
business plans and specifically, the system's ability to attract the 
ridership levels needed to avoid public operating subsidies. Local 
communities and property owners in California's Central Valley have 
also raised concerns about the project and its potential impact to the 
agriculture sector in the region. As we reported in our December 2012 
testimony, the Authority will face several challenges with acquiring 
rights-of-way in a timely manner, including potential construction 
delays as well as additional project costs.[Footnote 21] Timely right-
of-way acquisition will be critical since some properties are in 
priority construction zones. Property to be acquired will include 
homes, businesses, and farmland. Not having the needed right-of-way 
could cause delays and add to project costs. There are a total of 
approximately 1,100 parcels to be acquired for the first construction 
segment; all of which are in California's Central Valley. According to 
Authority officials, although the Authority may face challenges in 
acquiring right-of-way, they have built contingencies for time and 
cost into their acquisition plan. 

Authority Substantially or Partially Met GAO's Best Practices for 
Producing Reliable Cost Estimates, but Can Make Improvements: 

Reliable Cost Estimates Are Critical to Successful Project Planning, 
Funding, and Oversight, but Little FRA Guidance Is Available: 

The Authority estimates that Phase 1 of the high-speed rail project in 
California will cost $68.4 billion to construct and hundreds of 
millions of dollars to operate and maintain annually.[Footnote 22] 
Since the project's financing plan, as articulated in the April 2012 
revised business plan, will depend on an additional estimated nearly 
$38.7 billion in federal funds, it is vital that the Authority, FRA, 
and Congress be able to rely on these cost estimates for project 
planning, funding, and oversight. In addition, because the value of 
potential private investment depends on the cost of operating the 
system, it is vital that the Authority and the private sector be able 
to rely on the operating cost estimate. Given that our past work on 
high-speed rail projects around the world has shown that projects' 
cost estimates tend to be underestimated, ensuring the reliability of 
the estimates is critical to the success of this project.[Footnote 23] 

FRA provided limited guidance to grant applicants, including the 
Authority, about preparing cost estimates. FRA grant applicants were 
required to submit detailed capital cost estimates and high-level 
operating cost estimates, but FRA did not provide guidance on how 
applicants should produce these cost estimates to help ensure 
reliability. Moreover, the limited guidance that was provided did not 
reflect best practices included in our Cost Guide. FRA officials 
acknowledged that they specified the categories and types of costs to 
be estimated, not how applicants should prepare these cost estimates. 
FRA officials told us that they did not provide prescriptive guidance 
to grantees in preparing cost estimates because of the Recovery Act 
requirement to begin funding activities quickly following the 
enactment of the act in February 2009. In addition, FRA noted that the 
first two rounds of the HSIPR program were open to a wide range of 
project types and the level of detail necessary for an individual 
station is different from the level of detail necessary for a large, 
long-term corridor program like California. 

According to FRA officials, the Authority's application complied with 
the HSIPR grant application requirements. FRA found the cost estimates 
to be reasonable based on their comparison of the Authority's cost 
estimates (on a unit cost basis) to other rail projects in the United 
States and abroad. The Authority and its contractor told us that, in 
the absence of specific guidance on preparing cost estimates, they 
relied on their professional experience supplemented by available cost-
estimating guidance from the Federal Transit Administration (FTA) 
which they thought was the most applicable guidance available. For 
example, since FRA's guidance did not require the Authority to perform 
an independent cost estimate (this involves a comparison of the 
Authority's original cost estimates to those performed by an 
independent entity), the Authority turned to FTA guidance to provide 
direction on how and when to conduct an independent cost estimate. 

We evaluated the Authority's cost estimates against GAO's Cost Guide, 
which details best practices for generating high-quality cost 
estimates at all levels of government. While not required by FRA, the 
best practices identified in our Cost Guide help estimators develop 
reliable cost estimates, which have the four following characteristics: 

* An accurate cost estimate is unbiased, not overly conservative or 
overly optimistic, and based on an assessment of most likely costs. 

* A credible cost estimate discusses any limitations of the analysis 
from uncertainty or biases surrounding data or assumptions. 

* A comprehensive cost estimate ensures that costs are neither omitted 
nor double counted. 

* A well-documented cost estimate is thoroughly documented, including 
source data and significance, clearly detailed calculations and 
results, and explanations for choosing a particular method or 
reference. 

Ensuring that cost estimates reflect these four characteristics helps 
minimize the risk of cost overruns, missed deadlines, and unmet 
performance targets. The Cost Guide also provides criteria for 
evaluating cost estimates to determine whether they exhibit these 
characteristics. We have previously applied the Cost Guide in 
reviewing several transportation and infrastructure projects and we 
applied it in our review of the Authority's cost estimates.[Footnote 
24] 

Authority Substantially or Partially Met Best Practices in our Cost 
Guide for Producing Reliable Cost Estimates: 

The Authority substantially met best practices in our Cost Guide for 
producing accurate cost estimates, but only partially met our best 
practices for producing comprehensive, well documented, and credible 
estimates. By not following all best practices, there is increased 
risk of such things as cost overruns, missed deadlines, and unmet 
performance targets. 

Our assessment of the Authority's $68.4 billion construction and 
operating cost estimates for the high-speed rail project is summarized 
in table 1. Our assessment is discussed in more detail in appendix 2. 

Table 1: Summary of GAO's Assessment of California High-Speed Rail 
Authority's Cost Estimates: 

Characteristics of a reliable cost estimate: Accurate; 

Description of best practices: 
* Is based on most likely costs and is adjusted for inflation; 
* Is based on appropriate cost estimating techniques and contains 
minimal errors; 
* Reflects the current status of the project and has been updated with 
actual costs; 
* Is based on a historical record of cost estimating and comparable 
projects; 
GAO assessment: Substantially met. 

Characteristics of a reliable cost estimate: Comprehensive; 

Description of best practices: 
* Includes all project lifecycle costs such as construction, operation 
and maintenance; 
* Reflects the technical description of the project and is at an 
appropriate level of detail; 
* Reflects the ground-rules and assumptions of the project; 
GAO assessment: Partially met. 

Characteristics of a reliable cost estimate: Well documented; 

Description of best practices: Is supported by documentation that 
shows: 
* the source data and how they were adjusted; 
* the calculations and estimating methodology; 
* a description of how the estimate was created; 
* consistency with the technical baseline of the project; 
* review and approval from management; 
GAO assessment: Partially met. 

Characteristics of a reliable cost estimate: Credible; 

Description of best practices: Includes: 
* sensitivity analysis (assessing the impact of cost drivers); 
* risk and uncertainty analysis (identifying and quantifying risk); 
* cross-checks on major cost elements; 
* an independent cost estimate; 
GAO assessment: Partially met. 

Source: GAO. 

Note: GAO assessed the extent to which the Authority's project cost 
estimates met the best practices derived from the Cost Guide. The 
individual assessments are then used to determine an overall 
assessment for each characteristic. Assessments are made on a five-
point scale: Not Met, Minimally Met, Partially Met, Substantially Met, 
and Fully Met. See Appendix 1 for a description of how we conducted 
our assessment and Appendix 2 for more detail on the complete 
assessment. 

[End of table] 

Accurate: 

We found that the Authority substantially met best practices for 
developing accurate cost estimates. Consistent with best practices, 
the estimates reflect the new "blended" system, which will rely, in 
part, on existing rail infrastructure; they contained few, if any, 
mathematical errors; and they have been adjusted for inflation. 
Furthermore, the Authority's contractor used a construction industry 
database of project costs supplemented with actual bid-price data from 
other transportation infrastructure projects. While the Authority 
generally complied with best practices for producing accurate cost 
estimates, we could not determine whether the estimates were unbiased. 
This was the only best practice related to accuracy where the 
Authority fell short. To help ensure an unbiased estimate, the Cost 
Guide recommends conducting a systematic analysis of the potential 
risks to the project and their likelihood of occurring--called a risk 
and uncertainty analysis. A risk and uncertainty analysis is also a 
best practice for developing a credible cost estimate as discussed 
below. 

Comprehensive: 

We found that the Authority partially met best practices for producing 
comprehensive cost estimates. For example, the Authority met the best 
practice for including in the cost estimates the major components of 
the project's construction and operating costs. The construction cost 
estimate is based on detailed construction unit costs that are, in 
certain cases, more detailed than the cost categories required by FRA 
in its HSIPR grant application. However, the operating cost estimate 
was not as detailed as its construction cost estimate, as over half of 
the operating costs are captured in a single category called Train 
Operations and Maintenance. Authority officials told us that they 
developed their cost estimates consistent with FRA's guidance which 
emphasized greater detail on the construction cost estimate and less 
detail on the operating cost estimate.[Footnote 25] FRA officials 
confirmed that they emphasize construction cost estimates because 
HSIPR grants are required by federal law to only fund the capital 
costs of a project, not its operating costs. However, sufficiently 
comprehensive operating cost estimates are necessary to determine the 
potential profitability of California's project, a key consideration 
to attracting private-sector investment in the project that the 
Authority is counting on to help complete construction of the project. 
In addition, the Authority did not follow the best practice that calls 
for clearly describing certain assumptions underlying the construction 
and operating cost estimates. For example, Authority officials told us 
that the California project will rely on proven high-speed rail 
technology from systems in foreign countries, but it is not clear how 
the cost estimates were adjusted for applying the foreign technology 
in California and how these adjustments are reflected in the complete 
project cost estimates. California requirements on the technology may 
differ in terms of speed or safety requirements from that of other 
systems. Authority officials said that they produced a thorough 
description of the technical requirements for high-speed rail, but we 
were unable to see how this document was linked to the cost estimates, 
that is, a description of how these technical requirements will impact 
the project cost estimates. Without comprehensive cost estimates, it 
is not possible to independently assure that all cost-influencing 
factors and assumptions were considered. 

Well Documented: 

We found that the Authority partially met best practices for producing 
well-documented cost estimates. In many cases, the methodologies used 
to derive the construction cost estimate were well documented, but in 
other cases, were more limited. For example, while track 
infrastructure costs ($23.6 billion in 2011 dollars) were thoroughly 
documented, costs for other elements, such as building new stations 
and acquiring trains ($3.2 billion in 2011 dollars), were not 
supported with sufficient documentation to identify how these costs 
were developed and what costs were included or excluded. Authority 
officials told us that since station locations and train technology 
are not yet finalized, they used a higher-level cost estimate. 
Additionally, we were unable to trace the estimates back to their 
source data and recreate the estimates using the stated methodology. 
For example, we were unable to identify the basis for how the 
operating costs from analogous foreign high-speed rail projects were 
adjusted for use in California. Authority officials said that the 
operating cost estimate was used at a high level to determine whether 
or not the California system will operate with an operating surplus. 
Authority officials plan to refine the operating cost estimate as the 
project progresses. However, without more detailed documentation, the 
Authority's cost estimates are more difficult to support and it may be 
harder to make changes to the estimates as they are revised since the 
basis of the original estimate may not be documented. In addition, 
without more thorough documentation, FRA and other oversight officials 
cannot replicate and evaluate what the Authority did to prepare its 
estimates and potentially exposes the project to possible cost 
overruns because the basis for costs are not known. Estimates that 
lack documentation are not useful for updates or information sharing 
and can hinder understanding and proper use. 

Credible: 

We found that the Authority partially met best practices to help 
ensure the credibility of its cost estimates. Those practices include: 

* testing such estimates with a sensitivity analysis, such as 
assessing the effect of changes in key cost inputs; 

* obtaining an independent cost estimate conducted by an unaffiliated 
party to see how outside estimates compare to the original estimates; 
and: 

* conducting risk and uncertainty analysis. 

In regard to the construction cost estimates, the Authority performed 
a sensitivity analysis for the approximately first 30 miles of 
construction and obtained an independent cost estimate for the first 
185 miles of construction in the Central Valley but neither covered 
the entire Los Angeles to San Francisco project. And, as noted under 
the accuracy discussion, the Authority did not conduct a risk and 
uncertainty analysis on the cost estimates for any construction 
segment. Authority officials told us that in the absence of relevant 
FRA guidance, they followed FTA guidance for these types of 
evaluations. They noted that FTA guidance recommends conducting 
sensitivity tests once route alignments are selected and that, thus 
far, only 30 miles of the project meet those criteria. Similarly, 
Authority officials told us that they interpreted FTA guidance to 
require an independent cost estimate for those segments that have 
passed the 15 percent design milestone, which includes the first two 
construction segments from Merced to Fresno and Fresno to Bakersfield. 
Because the Authority's cost estimates cover both construction of the 
full system--from San Francisco to Los Angeles--and include operating 
costs, sensitivity analyses and independent cost estimates are more 
beneficial when they cover the entire project to help ensure greater 
credibility. The methodology of these tests can be altered to reflect 
the level of design, so a segment that has met a certain level of 
design can still be evaluated for credibility. Finally, as noted 
above, the Authority did not perform a risk and uncertainty analysis, 
which would improve the estimates' credibility by identifying a range 
of potential costs and indicating the degree of confidence decision-
makers can place on the cost estimates. For example, the Authority 
faces the potential challenge of acquiring rights-of-way in a timely 
manner. Authority officials told us there about 400 parcels in the 
first construction package, about 100 parcels of which are considered 
to be potentially at-risk for timely delivery for construction. 
However, without a risk and uncertainty analysis it is not possible to 
determine how the cost estimates might be affected by such things as 
delays in acquiring necessary rights-of-way or having to pay more for 
property to keep the project on schedule. 

Authority officials said that they did not conduct a risk and 
uncertainty analysis yet because FRA did not require one, and the FTA 
guidance to which they turned, recommends such an analysis after a 
final route alignment is selected. In addition, Authority officials 
told us that they added contingencies to the cost estimates to account 
for the risk of cost overruns. However, according to our Cost Guide, 
risk and uncertainty analysis--as with sensitivity analysis and 
independent cost estimates--should cover the entire cost estimate and 
can be performed at varying levels of detail commensurate to the level 
of design. And, while contingencies are designed to cover potential 
cost overruns, based on our review, the Authority's contingencies, 
which range from 10 to 25 percent for various cost elements, were not 
calculated based on the results of a risk and uncertainty analysis 
(since this was not performed) but rather were based on professional 
judgment. Without a risk and uncertainty analysis, we cannot be 
assured that the contingencies are accurately calculated, and more 
importantly, what level of confidence we can have in the cost 
estimates. 

For the operating cost estimate, the Authority conducted sensitivity 
tests under various ridership scenarios as recommended by the Panel; 
however, these tests were designed to measure the ability of the 
initial operating section to cover operating costs with ticket 
revenues and not to determine the potential risk factors that may 
affect the operating cost estimate itself. The Authority also did not 
compare its operating cost estimate to an independent cost estimate or 
conduct a risk and uncertainty analysis. The Authority told us that it 
views the sensitivity test already conducted as well as a forthcoming 
evaluation of operating costs by the International Union of Railways 
(UIC) as sufficient to meet these requirements. To make its operating 
cost estimate more comprehensive and better documented, the Authority 
has contracted with the UIC to evaluate the existing methodology and 
data and help refine the Authority's estimates.[Footnote 26] While the 
UIC's evaluation will likely provide an expert review of the 
Authority's operating cost estimates, it may not address some of the 
key practices that ensure credibility. For example, the UIC's 
evaluation is not expected to result in new, independently-produced 
cost estimates that can be compared to the Authority's original 
estimates. 

Authority Can Make Improvements to Its Cost Estimates as the Project 
Progresses: 

The quality of any cost estimate can be improved as more information 
becomes available. And, based in part on evaluations from the PRG, the 
Authority is taking some steps to improve the cost estimates that will 
be provided in the 2014 business plan. As noted above, the Authority 
has contracted with the UIC to evaluate and provide recommendations on 
the Authority's operating cost estimates. While the study will provide 
additional analysis from a reputable source, it may not address all 
best practices from the Cost Guide that would help ensure that the 
operating cost estimate is comprehensive, accurate and credible. 

Cost estimates should also be updated with actual costs so that they 
are always relevant and current. Continual updating of a cost estimate 
as a program matures not only results in a higher-quality estimate but 
also provides an opportunity to incorporate lessons learned. While the 
Authority was not able to incorporate actual costs because 
construction had not yet begun, it will have the opportunity once 
contracts are awarded and actual costs begin to incur for the initial 
construction in the Central Valley, which is expected to begin in 
2013. The bids for the first 30-mile construction package have been 
submitted to the Authority and will provide a check on how well the 
Authority has estimated the costs for this work, as well as provide 
more information on its cost estimates for other segments of the 
project.[Footnote 27] 

Ridership and Revenue Forecasts Are Reasonable for Current Purposes, 
but Will Require Updates: 

The Authority's ridership and revenue forecasts to date are reasonable 
and the methods used to develop them followed generally accepted 
travel-demand-modeling practices. In addition, the Authority completed 
several updates to its ridership-and-revenue forecasting model after 
the release of the April 2012 revised business plan and also completed 
several sensitivity analyses to test the reasonableness of its model. 
However, the Authority will need to complete several additional 
updates to improve its model and the resulting forecasts for the 2014 
business plan. Authority officials stated that they have plans in 
place to complete several critical updates, including completing a new 
travel preference survey and developing a second generation travel 
demand model, but will not be able to complete these improvements in 
time for the 2014 business plan. 

Forecasts Are Reasonable Based on Use of Accepted Travel-Demand-
Modeling Practices, but Ridership and Revenue Model Will Require 
Additional Updates: 

Based on our review, we found that the Authority's methods and model 
used to produce its ridership and revenue forecasts adhere to 
generally accepted travel-demand-modeling practices. However, the 
Authority will need to complete several updates to improve these 
forecasts for the 2014 business plan. In its April 2012 revised 
business plan, the Authority forecasts between 16.1 million and 26.8 
million passengers per year and annual revenues of $1.06 billion to 
$1.81 billion in 2030. These forecasts were derived from a statewide 
ridership and revenue model, developed under contract to the 
Metropolitan Transportation Commission (MTC).[Footnote 28] Developing 
travel demand and revenue forecasts is difficult in almost every 
circumstance, across every mode and for a variety of reasons. As we 
have previously reported, limited data and information, especially 
early in a project before specific service characteristics are known, 
make developing reliable ridership and revenue forecasts difficult. 
[Footnote 29] To the extent early stage data and information are 
available they need to be updated to reflect changes in the economy, 
project scope, and consumer preferences. In addition, risks of 
inaccurate forecasts are a recurring challenge for project sponsors. 
Research on ridership and revenue forecasts for rail infrastructure 
projects around the world have shown that ridership forecasts are 
often overestimated. Although forecasting is inherently risky, 
reliable ridership and revenue forecasts are critical to accurately 
estimate the financial viability of a high-speed rail project and 
determine what project modifications, if any, may be needed. Such 
forecasts enable policymakers and private entities to make informed 
decisions about the proposed project and to determine the associated 
risks when making investment decisions. In addition, ridership 
forecasts are critical because they serve as the basis for revenue 
forecasts. If the California high-speed rail project is unable to 
generate the necessary ridership and revenue to cover the system's 
operating costs, the project may not be able to operate without a 
subsidy--as required by Proposition 1A. Conversely, if forecasts are 
overly conservative, it could lead to the state capturing less value 
from private investment than warranted. As such it is critical that 
the Authority's process for developing these forecasts is reliable and 
provides some assurance that the resulting forecasts provide a 
reasonable estimate of future demand for the system. 

Unlike our cost estimating criteria discussed earlier, there is no 
industry standard or established criteria for developing or evaluating 
intercity passenger high-speed rail ridership forecasts. FRA has not 
established guidance on acceptable approaches to the development of 
reliable ridership and revenue forecasts and has established only 
minimal requirements and guidance related to information HSIPR grant 
applicants must provide regarding forecasts. We previously reported 
that developing guidelines, methods, and analytical tools to develop 
credible and reliable ridership forecasts is necessary to ensure 
equitable consideration of high-speed rail as a potential option to 
address demands on the nation's transportation system.[Footnote 30] We 
recommended that the Secretary of Transportation develop guidance and 
methods for ensuring the reliability of ridership and other forecasts 
used to determine the viability of high-speed rail projects and 
support the need for federal grant assistance.[Footnote 31] The DOT 
OIG has also recommended that FRA develop specific and detailed 
guidance for the preparation of HSIPR ridership and revenue 
forecasts.[Footnote 32] According to FRA officials, they are in the 
process of developing an oversight plan that will include criteria for 
evaluating ridership forecasts. FRA officials indicated that they 
intend to use the DOT OIG's HSIPR Best Practices: Ridership and 
Revenue Forecasting guide as a starting point for developing this 
guidance.[Footnote 33] 

For the purposes of our assessment, we identified generally accepted 
travel-demand-modeling practices for high-speed rail projects from a 
variety of sources and developed criteria based on these practices to 
assess the reasonableness of the approach used to create the ridership 
and revenue models for the California high-speed rail project. 
[Footnote 34] In developing our criteria, we relied primarily on a 
2011 report, prepared for the DOT OIG's office by the firm Steer 
Davies Gleave, on best practices related to travel demand modeling. 
[Footnote 35] In addition, we also examined other literature on 
developing rail ridership and revenue forecasts to supplement 
information in the Steer Davies Gleave report. Specifically, we 
reviewed, among other sources, our prior GAO reports, Federal Highway 
Administration (FHWA) and FRA guidance, and academic literature. (See 
appendix I for a list of guidance and reports we reviewed.) From our 
review of these reports and other sources, we identified common 
approaches to developing ridership and revenue forecast models and 
elements affecting the validity of those models. We identified seven 
key steps of the ridership-and-revenue forecasting process and then 
compared the Authority's process for completing these tasks to 
generally accepted travel-demand-modeling practices. We found that the 
Authority followed generally accepted travel-demand-modeling practices 
for each of the seven key steps: (1) developing trip tables, (2) 
determining and applying service characteristics, (3) developing mode 
choice models, (4) estimating induced travel, (5) estimating expected 
fare revenue, (6) conducting sensitivity analysis, and (7) conducting 
validation testing. (For a detailed description of generally accepted 
travel-demand-modeling practices and the Authority's process for 
completing each of these steps, see appendix III). 

Developing Trip Tables: 

The Authority's process for developing trip tables and collecting and 
compiling data on current travel patterns followed generally accepted 
practices. A central task of the ridership forecasting process 
involves collecting and compiling data on current travel patterns 
along the proposed high-speed rail route into trip tables.[Footnote 
36] We found that the Authority followed generally accepted practices 
for developing trip tables. For example, trips were distinguished by 
mode of travel (auto, air, rail), trip purpose (commute, business, 
recreation, and other), and trip length (over and under 100 miles) as 
is general practice. Various data sources were used to develop the 
base year trip tables, including, among other sources, 2000 Census 
Bureau data, survey data from a 2005 travel survey[Footnote 37] and a 
2011 long-distance travel survey conducted by Harris Interactive, 
[Footnote 38] and existing regional models used by metropolitan 
planning organizations (MPOs).[Footnote 39] Overall, the data sources 
used for developing the trip tables were consistent with generally 
accepted standards. One potential limitation, however, is that the 
2011 survey sample was not selected at random from among California 
residents but rather was limited to individuals who had opted to join 
an online survey panel. The American Association for Public Opinion 
Research (AAPOR)[Footnote 40] has recommended against the use of such 
panels, often called "opt-in" survey panels, when accurate population 
estimates are needed, due to concerns about data quality and the 
possibility that a panel may differ from the intended target 
population in unknown ways. Data were weighted to adjust for 
differences between the survey sample and the California population on 
four characteristics--geographic location, age, wealth, and employment 
status--but, the data may still not be representative of the 
California population in other characteristics related to travel 
behavior. 

Determining and Applying Service Characteristics: 

The Authority's process for determining and applying service 
characteristics followed generally accepted practices; however, as 
detailed service plans are finalized or scenarios are changed, the 
model will need to be updated to reflect revised service 
characteristics. Ridership forecasts require information on the 
service characteristics (such as travel time and fares) of competing 
modes of travel, such as automobile and air travel along the proposed 
route. We found that the Authority collected and considered relevant 
service characteristics and used appropriate data sources. These 
included information on service characteristics of the interregional 
transportation modes in the area including information on time, cost, 
and other service characteristics for each mode--auto, conventional 
rail, and air--based on the most currently available published or 
observed data. High-speed rail service characteristics were defined 
based on the initial service plans and fare structure because 
published or observed data do not exist. Some or all of the high-speed 
rail characteristics will likely change as service plans are finalized 
and engineering decisions are made, and those changes can 
significantly affect ridership and revenue forecasts. This was 
illustrated in a 2012 sensitivity analysis completed by the Authority 
in which service characteristics for the proposed high-speed rail 
system were adjusted to reflect reduced service on the San Francisco 
Peninsula, which is likely under the current "blended" approach 
whereby the high-speed rail system will share tracks with Caltrain. 
[Footnote 41] This adjustment reduced high-speed rail ridership and 
revenue forecasts by 11 percent and 13 percent, respectively, compared 
with forecasts in the April 2012 revised business plan. Updated 
representations of the base and forecast year level of service 
characteristics will be important for producing realistic ridership 
forecasts in the future. 

Developing a Mode Choice Model: 

While the Authority mostly followed generally accepted practices as 
well as used some more advanced techniques in its mode choice model 
development, peer and academic reviewers such as the California High-
Speed Rail PRG, the Panel, and academic experts from the University of 
California Berkeley's Institute of Transportation Studies (ITS), 
[Footnote 42] have identified limitations with the survey data used to 
develop the mode choice model. "Mode choice models"--or travel 
behavior models--illustrate how different variables such as income, 
travel time, and travel costs affect travel choices, including 
decisions about which mode of transportation to use. The reliability 
of the mode choice model estimates depends heavily on the quality of 
data used, which in turn depends on methods used to collect the data. 
[Footnote 43] These models can be based on traveler behavior observed 
in actual travel situations or self reported via a survey (revealed 
preference data), or on hypothetical situations presented to travelers 
in a survey (stated preference), or both. 

There are advantages and disadvantages to using revealed preference 
and stated preference data. Revealed preference data provides 
information on travelers' actual choices made in a specific market. 
However, according to the Steer Davies Gleave report, when collected 
from travel surveys, biases may exist in the respondents' responses 
due to a desire of respondents to justify their chosen mode. In 
addition, since true high-speed rail does not exist in the U.S., it is 
not possible to use revealed preference alone to determine how 
American travelers would actually use high-speed rail. To address this 
problem, stated preference data have been used in high-speed rail 
studies to asses likely traveler responses to a new service. According 
to the Steer Davies Gleave report, while stated preference data can 
provide detailed information about a traveler's likely responses to 
different modes or services that do not currently exist, this type of 
data may also exhibit bias; specifically, survey respondents may 
respond favorably to a hypothetical new mode, when in reality it may 
be more difficult to change habitual behavior. According to the Steer 
Davies Gleave report, limitations with each of these types of data can 
be mitigated by taking various steps such as by combining both 
revealed preference and stated preference data. The primary source of 
data for the Authority's mode choice model was a revealed preference 
and stated preference survey, of air, rail, and auto trip passengers, 
conducted at airports, rail stations, and by telephone from August to 
November 2005. However, in its July 2011 and May 2012 reports to the 
Authority, the Panel reported that the Authority's main mode choice 
model was based solely on stated preference responses, and recommended 
in its May 2012 report that the Authority collect new survey data and 
use both revealed and stated preference data in developing a new mode 
choice model. Authority officials stated that they are currently 
developing a new revealed preference and stated preference survey, 
which they plan to begin administering in early 2013. We discuss this 
survey in further detail in the next section. 

In addition, academic experts from the University of California 
Berkeley's ITS[Footnote 44] have previously reviewed the ridership and 
revenue model and have also identified limitations in the Authority's 
method of applying a statistical model to the survey data. These data 
came from a choice-based sample, meaning that survey respondents were 
selected for having already chosen a mode of transport (air, rail, or 
auto). The Authority used a conventional method of applying a 
statistical model to such data, but a newer method has been identified 
in a recent research paper.[Footnote 45] According to some academic 
experts, the use of the latest method could be an improvement. The 
Panel also reviewed this issue in its May 2012 report and stated that 
while they do not see the non-use of this new method in the Version 1 
model as an important defect, it is worth investigating this issue as 
the Authority continues to refine the travel demand model. 

Estimating Induced Travel: 

The Authority followed generally accepted practices when producing 
induced travel estimates. Induced travel refers to trips that occur as 
a result of the high-speed rail project and that might not otherwise 
have been made using existing modes. In general, these are new trips 
that are generated because a new travel mode exists. The Authority 
estimated induced travel for the California high-speed rail project to 
be on average 2 percent of total high-speed rail trips. According to 
the Steer Davies Gleave report, based on its review of forecasts of 
proposed U.S. high-speed rail systems, an upper limit on induced 
travel of approximately 10 percent of total high-speed rail trips is 
widely accepted for proposed U.S. high-speed rail systems. Steer 
Davies Gleave also reviewed actual induced travel for high-speed rail 
systems outside the United States and found that it ranged from 6 to 
27 percent. The Authority forecasted induced travel to be on average 2 
percent. This estimate appears to be conservative and reasonable when 
compared to proposed U.S. high-speed rail systems and actual induced 
travel for high-speed rail systems outside the United States. 

Estimating Expected Fare Revenue: 

The Authority followed generally accepted practices when estimating 
expected fare revenue for the California high-speed rail project. 
Expected fare revenue is a product of forecast ridership and average 
fares. High-speed rail fares are based on a boarding fare and a per 
mile fare for interregional trips. For example, travel from Los 
Angeles to San Francisco would mean an average, one-way, high-speed 
rail fare of $81 in 2010 dollars or 83 percent of average 2009 
airfares from Los Angeles to San Francisco. Forecast ridership is 
impacted by high-speed rail fares and revenue is an output of the 
ridership and revenue model. According to the Authority, the annual 
number of riders needed to break-even when the Phase 1 blended system 
is opened in 2029 is 6.1 million or 23 percent of the high forecast. 
Authority officials stated that they did not produce a revenue 
optimization forecast--that is, a ridership forecast that would 
maximize revenue--to produce these high-speed-rail fare estimates and 
acknowledged that they will need to do so in the future when meeting 
with potential private high-speed rail operators, which will establish 
their own revenue-maximizing fares for the system. According to the 
Authority, the Authority will not operate the high-speed rail system 
and a private operator is expected to serve as a contract operator of 
the system. As such, the private operator will be expected to assume 
all revenue risks of the project (including setting fares). The 
Authority's process for calculating expected fare revenue adheres to 
generally accepted practices; however, as other factors and inputs 
change, total expected fare revenue will likely also change. For 
example, over time, it will be important for the Authority to monitor 
changes in airfares, gasoline prices, and other key assumptions and 
incorporate these changes, as necessary, into future revenue forecasts. 

Conducting Sensitivity Analysis: 

The Authority followed generally accepted practices when conducting 
sensitivity analysis of key model assumptions for the California high-
speed rail project. A sensitivity analysis is typically conducted by 
varying key model assumptions such as socioeconomic data, type of 
trips taken, gasoline and auto fleet efficiency, and airfares, or 
parameter values to determine how the model behaves in response to 
changes to these assumptions. The Authority conducted several 
sensitivity analyses on its ridership and revenue model, most in 2011. 
For example, the Authority conducted sensitivity analyses that tested 
key factors, such as changes in fuel economy, air and auto travel 
time, air and auto travel costs, and high-speed rail travel time 
assumptions. In one analysis, the Authority tested the overall effect 
of a higher auto fuel economy and found that this change reduced 
ridership and revenue forecasts by 16 and 19 percent, respectively, 
from the Phase 1 high ridership and revenue forecasts presented in the 
November 2011 draft business plan. According to the Panel that 
reviewed these analyses, the results of the various sensitivity 
analyses that the Authority has conducted show that the model is 
appropriately sensitive across the range of variables tested. 

In addition, the Authority performed a sensitivity analysis of an 
extreme downside scenario to test the ridership and revenue 
implications of a series of downside events, such as increased average 
rail travel time from Merced to the San Fernando Valley (140 minutes 
instead of 126 minutes), decreased train frequency (3 trains per hour 
instead of 4 trains per hour during peak times), lower auto-operating 
costs, and lower air fares (10 percent below actual 2009 average air 
fares). Based on this analysis, the Authority determined that an 
extreme downside scenario would be expected to reduce ridership and 
revenue forecasts by 27 percent and 28 percent, respectively, below 
that shown for the low IOS forecasts in the April 2012 revised 
business plan. According to the Authority, they tested these events 
using the IOS phase because the financial viability will be the most 
fragile during the early stages of ridership. According to the 
Authority, these forecasts are still sufficient to cover the 
Authority's estimated operating costs and not require a public 
operating subsidy. Authority officials stated that they intend to 
conduct additional sensitivity analyses going forward. 

Conducting Validation Testing: 

The Authority mostly followed generally accepted practices to validate 
the ridership and revenue model. Model validation generally involves 
verifying that the model reflects observed traveler behavior including 
total travel, region-to-region travel flows, and observed market 
shares by mode. In the United States, without another high-speed rail 
system to use for comparison, model validation is a difficult task. 
Furthermore, validating the proposed California project with foreign 
high-speed rail systems is difficult because some of the travel market 
characteristics in other countries with high-speed rail, such as the 
cost of driving, may not be comparable. The Authority has taken some 
steps to validate the model through tests performed using data on 
Amtrak's premium Acela service in the Northeast Corridor (NEC) as 
input to the California high-speed rail model and compared the output 
with 2008 actual ridership and 2030 NEC forecasts.[Footnote 46] The 
results of the Authority's comparison of the California high-speed 
rail model with NEC input to actual 2008 ridership data showed that 
ridership in the California high-speed rail model with NEC-like 
conditions in 2008 is 79 percent of actual 2008 NEC ridership. 
Similarly, the Authority's comparison of the California high-speed 
rail system ridership forecasts from the model run with "NEC-like" 
service is about the same as projected 2030 ridership on the Acela 
service. Authority officials told us they believe the results from 
these tests demonstrate that the ridership and revenue model is 
reasonably sensitive to speed, frequency, and fares. While the NEC 
corridor has some comparable characteristics to the proposed 
California high-speed rail corridor--such that it is capable of 
reaching top speeds up to 150 miles per hour and covers a distance of 
over 400 miles--the proposed California high-speed rail corridor and 
the NEC differ in important ways. For example, population density, 
[Footnote 47] congestion, and travel behavior in the two corridors 
differ, and as such, forecast comparisons should be interpreted with 
caution. 

In developing the forecasts for the April 2012 revised business plan, 
the Authority also revised several model assumptions used in the 
initial ridership and revenue forecasts presented in the November 2011 
draft business plan. Specifically, the Authority revised model 
assumptions to reflect changes in current and anticipated future 
conditions for airfares and airline service frequencies, decreases in 
gasoline price forecasts, and anticipated declines in the growth rates 
for population, number of households, and employment. Some of the 
initial assumptions were largely based on pre-2007 data and did not 
reflect potential effects of the 2007 to 2009 recession. (See table 2 
for a summary of updates that have thus far been completed.) According 
to Authority officials, this was done to build in additional 
conservatism in the ridership forecasts. 

Table 2: Description of Updated California High-Speed Rail Travel-
Demand-Model Assumptions: 

Model input assumptions: Airfares and airline service frequencies; 

Description of revised assumptions: 
* Authority updated the model to reflect changes in current and 
anticipated future airfares and airline service frequencies. 
Initially, the model reflected 2005 average fares and airline service 
frequencies--it was updated to reflect 2009 average fares and airline 
service frequencies; 
* The Authority engaged Aviation System Consulting, LLC (ASC), a 
California-based expert firm to address these issues. ASC analyzed the 
past decade of U.S. Department of Transportation data on airline 
service and fare levels, explained the economic factors affecting 
airline responses to changes in competition and capacity, and helped 
determine scenarios of potential airline competitive response to the 
introduction of high-speed rail service. 

Model input assumptions: Conventional rail service and fare 
assumptions; 

Description of revised assumptions: 
* Conventional rail service and fare assumptions for 2030 were updated 
to reflect 2011 current and forecasted conditions; 
* Authority updated the fare assumptions for all lines to on-line 
published fares in 2011. 

Model input assumptions: Revised automobile operating costs 
assumptions; 

Description of revised assumptions: 
* Gasoline price forecasts increased: The Authority adjusted its 
assumptions related to automobile operating costs to reflect the most 
recent forecasts from the U.S. Energy Information Administration, 
adjusted to reflect the historical difference between California 
prices and the national average; 
* Vehicle fuel economy forecasts: To align with Corporate Average Fuel 
Economy standards, forecasts were revised to show an average of 3 
percent to 6 percent fuel economy growth; 
* 2030 auto operating forecasts: The Authority estimated a range for 
auto operating costs incorporating both fuel and nonfuel components. 
The Authority revised 2030 operating costs for the 2012 revised 
business plan. The Authority used $0.20 per mile in 2011 dollars for 
the low model scenario and $0.28 per mile for the high model scenario. 

Model input assumptions: Socioeconomic forecast updates; 

Description of revised assumptions: The 2030 population, household, 
and employment forecasts used for the original ridership and revenue 
model were developed in 2006-2007 from local agency socioeconomic 
projections. The Authority developed two alternate forecasts of 
population, households, and employment to account for decreased 
expectations regarding future socioeconomic growth resulting from the 
2007-2009 recession: 
* High forecast: To develop the 2030 forecasts, the Authority 
purchased forecast data of county-level population, households and 
total employment for the state, for two years--2008 (pre-recession 
data) and 2011 (post-recession data). A combination of these data were 
used to account for pre-and post-recession impacts to factor 2030 pre-
recession forecasts; 
* Low forecast: 2011 county-level forecasts of population, households, 
and employment by economic sector for the state were purchased and 
used to develop 2030 forecasts. 

Model input assumptions: Alternative trip frequency assumptions; 

Description of revised assumptions: 
* The Authority developed a long-distance travel survey to collect 
current long-distance travel data to help provide perspective for the 
2012 revised business plan; 
* The Authority contracted with Harris Interactive for the data 
collection, and the survey was fielded in May 2011. 

Source: California High-Speed Rail Authority information. 

[End of table] 

Updating model assumptions can help mitigate of the risks of 
overestimating ridership and revenue forecasts--referred to as 
optimism bias. Biased ridership forecasts are a recurring problem with 
rail infrastructure projects and we have previously reported that 
forecasting ridership and revenue is a complex and uncertain process 
and ridership forecasts of high-speed rail projects are often 
overestimated.[Footnote 48] Other research on ridership and revenue 
forecasts for rail infrastructure projects have confirmed that actual 
ridership is likely to be lower than forecasted unless steps are taken 
to incorporate more conservative assumptions into the model. For 
example, a recent study examined a sample of 62 rail projects and 
found the ridership forecasts of 53 of them were overstated; actual 
ridership was, on average, 41 percent lower than forecasted.[Footnote 
49] Updates to model inputs, such as fuel prices and other 
projections, are important for updating ridership forecasts for any 
project; in this instance, updates to model inputs resulted in more 
conservative ridership forecasts. 

Next Business Plan Will Not Reflect All Planned Improvements to 
Ridership Forecasts: 

The Authority has plans to complete future improvements to its 
ridership and revenue forecasts, including completing a new travel 
preference survey and developing a second generation travel demand 
model. However, the Authority will not be able to complete these 
critical improvements in time for the 2014 business plan. According to 
Authority officials, the 2014 plan is expected to include, among other 
updates, updated ridership and revenue forecasts. Two critical updates 
to the ridership forecasts that peer reviewers and academic experts 
have recommended are the development of a new 2013 revealed and stated 
preference travel survey and a second generation travel demand model--
which will make use of the new survey. Although the Authority has 
begun taking steps to complete both of these tasks, neither the survey 
nor the second generation model will be completed in time for the 2014 
business plan. 

The Authority began work developing a new 2013 revealed and stated 
preference survey in late 2012. According to Authority officials, a 
survey sample and survey questionnaire design was initiated in 
December 2012. Data collection for the 2013 revealed preference and 
stated preference survey will not begin until 2013. Full data 
collection, cleaning, and preliminary analysis of results are expected 
to be completed by mid-April 2013. The new survey will include a 
larger sample of 4,500 respondents compared with the 3,172 respondents 
to the 2005 survey. In addition, the revealed preference portion of 
the data set will be designed and coded to facilitate estimation using 
revealed and stated preference responses simultaneously, which was not 
done in the first version of the model. According to the Panel, 
development of a new survey is critical as it will address several 
long-term issues that can only be overcome with the collection and 
analysis of new survey data. 

The Authority also has plans to develop a second-generation travel 
demand model; however, the second generation model will not be 
completed in time for the 2014 business plan. According to Authority 
officials, work on the second-generation model will not begin until 
the ridership and revenue analysis for the 2014 business plan is 
completed. According to the Authority, the second generation model 
will use data from the new 2013 revealed and stated preference survey 
to supplement data from the 2005 survey. In addition, the Authority 
plans to replace the 2011 Harris Interactive long-distance travel 
survey data with data from the 2012 California Household Travel Survey 
being conducted by the California Department of Transportation. 
According to Authority officials, both surveys will be needed for 
developing its second-generation travel demand model. The Panel, which 
has released five reports assessing the Authority's ridership and 
revenue forecasts, has reported that the Authority's ridership and 
forecasts to date are reasonable for planning purposes but has also 
stated that additional updates and enhancements, particularly the 
development of a new model, will be critical for future project 
decision making. For example, in its most recent October 2012 report, 
the Panel stated that a second-generation model will be required to 
meet the Authority's long-term goals of completing detailed planning 
studies and make key planning and operational decisions on issues such 
as specific rail alignments, station design requirements, and pricing 
strategies. 

While the Authority will not be able to complete a second-generation 
travel demand model in time for the 2014 business plan, it has begun 
work on developing an enhanced model that will be used to produce the 
ridership and revenue forecasts for the 2014 business plan. The 
enhanced model will retain the same structure of the original model, 
but some of the individual model components will be updated. For 
example, the main mode choice model will use both revealed preference 
and stated preference results from the 2005 travel survey. According 
to Authority officials, the enhanced travel demand model will be 
completed by May 31, 2013. 

Even if the Authority is not able to complete the major ridership and 
revenue forecast improvements in time for the 2014 business plan, 
ongoing disclosure of interim results from model improvements both 
before and after the business plan are published will be important to 
outside reviewers and the public. Peer reviewers and other groups that 
have examined the Authority's ridership and revenue forecasts have 
reported the need for greater transparency in the Authority's 
analyses. For example, in its January 2012 report, the California 
State Auditor reported that the Authority's November 2011 draft 
business plan lacked detail in its presentation of some of the revenue 
forecasts. Similarly, in its October 2012 report, the Panel advised 
the Authority to provide summaries on the Authority's website, 
describing, among other things, recent forecasts, key input 
assumptions used to develop the forecasts (e.g., fuel price trends, 
socioeconomic growth rates, and changes in household size and 
structure), and updated service characteristic information. According 
to Authority officials, documentation supporting analysis in its 2014 
draft business plan will be available on the Authority's website when 
the plan is made public. In addition, Authority officials stated that 
the Authority has generally posted key technical documents as they are 
made public. 

The Authority also has developed a work plan for other travel demand 
model improvements. This includes, among other things, plans to 
complete additional validation testing of model results using data 
from the NEC. For the 2014 business plan, the Authority is planning to 
conduct further testing and sensitivity analysis of the ridership and 
revenue forecasts to examine the sensitivity of the forecasts to 
reduced frequencies of services; changes in alternative fare 
structures (for example, premium fares for intra-regional trips in the 
San Francisco Bay Area and Los Angeles Basin); changes to service 
plans (destinations and schedules); and other sensitivity analyses 
aimed at quantifying risks. In its October 2012 report, the Panel also 
recommended additional sensitivity analyses to be completed on the 
second generation travel demand model, including analyses examining 
the impact of pricing strategies on revenue, impact of local transit 
feeder systems on station choice, and impact of major changes in the 
roadway network on highway congestion and subsequent mode choice 
decisions. In addition, the Authority is planning to conduct Monte 
Carlo simulations to test numerous potential combinations of 
assumptions on the forecasts that will be part of the 2014 Business 
Plan during fiscal years 2013 through 2014, provided the foundational 
information to construct, test, and analyze the simulation and its 
results is sufficiently developed at that stage of the program. All of 
these will be important as the ridership and revenue forecasts 
continue to evolve with development of the high-speed rail project. 

The Authority's Funding Plan Faces Uncertainty: 

The project's funding, which relies on both public and private sources 
of financing, faces uncertainty about whether those funds can be 
obtained in a tight federal and state budget environment. The 
Secretary of Transportation and the Governor of California have 
committed to funding this project, but obtaining sustained 
congressional and public support for appropriating additional funds is 
one of the biggest challenges to completing this project. In the 
latter stages, the Authority will also rely on private-sector 
financing, but will require more reliable operating cost estimates and 
revenue forecasts to determine whether, and the extent to which, the 
system will be profitable, as well as the value of any private 
investment. The Authority's financing plan recognizes the uncertainty 
of the current funding environment so the Authority is building the 
project in phases and has identified an alternative funding source 
that is also uncertain. However, delays in obtaining funds as planned 
will likely lead to project delays and higher costs for construction. 
A summary of funding already committed to date can be found in table 3. 

Table 3: Funding Secured for Constructing the California High-Speed 
Rail Project, as of February 2013: 

State high-speed rail bonds; $8.2 billion[A]. 

Federal HSIPR grants; $3.3 billion[B]. 

Total secured funding; $11.5 billion. 

Source: GAO analysis of FRA grant information and the California High-
Speed Rail Authority's April 2012 Revised Business Plan. 

[A] The Authority expects approximately $8.2 billion in proceeds from 
the $9.95 billion authorized in Proposition 1A high-speed rail bonds 
to be available for construction of high-speed rail. The remainder is 
for connectivity projects and engineering and environmental work. 

[B] Approximately $3.3 billion of $3.5 billion in obligated HSIPR 
grants is available for construction of high-speed rail project. The 
remainder is for engineering and environmental work. 

[End of table] 

Additional Public Sector Funding Is Needed to Complete the Project, 
but Securing It Will Be Challenging: 

The Authority is relying primarily on public-sector funding to 
complete construction of Phase 1 of the project with $55 billion, or 
81 percent of the total construction cost, expected to come from state 
and federal sources. Heavy reliance on public-sector funding is not 
unusual for a project of this size. For example, France built its high-
speed train system primarily with public-sector funding. In the United 
States, federal-aid highway system continues to be funded by the 
federal government through the gas tax and, more recently, with 
transfers from the general fund.[Footnote 50] The Authority expects to 
obtain public sector funds over the life of the project as individual 
segments are ready for construction. This type of "phased" funding is 
typical for major transportation infrastructure projects. Table 4 
provides a summary of the Authority's funding plan for Phase 1 the 
high-speed rail project. 

Table 4: The Authority's Funding Plan for Constructing Phase 1 of the 
California High-Speed Rail Project, as of April 2012: 

Funding source: Federal; 
First construction: $3.3 billion; 
Initial operating segment: $20.3 billion; 
Bay-to-Basin: $8.4 billion; 
Phase 1 blended: $10.0 billion; 
Total: $42.0 billion; (61%). 

Funding source: State high-speed rail bond; 
First construction: $2.7 billion; 
Initial operating segment: $4.4 billion; 
Bay-to-Basin: 0.0; 
Phase 1 blended: $1.1 billion; 
Total: $8.2 billion; (12%). 

Funding source: Locally generated; 
First construction: 0.0; 
Initial operating segment: $0.7 billion; 
Bay-to-Basin: $1.2 billion; 
Phase 1 blended: $3.1 billion; 
Total: $5.0 billion; (7%). 

Funding source: Subtotal public; 
First construction: $6.0 billion; 
Initial operating segment: $25.4 billion; 
Bay-to-Basin: $9.6 billion; 
Phase 1 blended: $14.2 billion; 
Total: $55.2 billion; (81%). 

Funding source: Private investment; 
First construction: 0.0; 
Initial operating segment: 0.0; 
Bay-to-Basin: $10.1 billion; 
Phase 1 blended: $3.0 billion; 
Total: $13.1 billion; (19%). 

Funding source: Operating cash flow; 
First construction: 0.0; 
Initial operating segment: 0.0; 
Bay-to-Basin: $0.2 billion; 
Phase 1 blended: 0.0; 
Total: $0.2 billion; (0). 

Funding source: Subtotal private investment and operating cash flow; 
First construction: 0.0; 
Initial operating segment: 0.0; 
Bay-to-Basin: $10.3 billion; 
Phase 1 blended: $3.0 billion; 
Total: $13.3 billion; (19%). 

Funding source: Total; 
First construction: $6.0 billion; 
Initial operating segment: $25.4 billion; 
Bay-to-Basin: $19.9 billion; 
Phase 1 blended: $17.2 billion; 
Total: $68.5 billion (100%). 

Source: GAO analysis of California High-Speed Rail Authority's April 
2012 revised business plan. 

[End of table] 

Federal Funding: 

The Authority's April 2012 revised business plan relies on 
approximately $42 billion in federal funding for the project's 
construction, which includes the $3.3 billion that has already been 
obligated. The remaining $38.7 billion in federal funds have not been 
identified in federal budgets or appropriations but would amount to an 
average of more than $2.5 billion annually over the life of the 
project's construction. This exceeds the average annual funding made 
available under DOT's New Starts transit-funding grant program since 
2008 (about $1.6 billion per year). Moreover, it exceeds the federal 
government's average annual appropriations to Amtrak since 2008 (about 
$1.5 billion per year). Largely as a result of the funding challenge, 
the Authority is taking a phased approach--planning to build segments 
as funding is available. Thus, according to the Authority's 2012 
revised business plan, no additional funding will be needed until 2015 
when it hopes to begin construction beyond the first construction 
segment. 

Based on our past work on high-speed rail, successful projects require 
significant and sustained financial commitments from the public sector 
before private investors will participate, and the Authority's plan 
reflects this funding model.[Footnote 51] For example, in Japan, 
private investment is contingent on substantial government investment. 
Other federally-supported transportation programs--like those for 
highway and certain transit infrastructure--rely on a dedicated 
revenue source for their funding and allow for multi-year funding 
agreements for eligible projects. In contrast, the HSIPR program has 
not been funded with a dedicated revenue stream, but from the general 
fund, a process that means that the program has to compete for 
appropriations with other discretionary programs.[Footnote 52] In 
addition, the HSIPR program has provided one-time grants and, as 
currently structured has not awarded multi-year agreements for 
grantees. Our 2013 High-Risk Series report identified the use of 
general funds for high-speed rail projects as a challenge to project 
completion.[Footnote 53] Given that the HSIPR grant program has not 
received funding since 2010 and that future funding proposals will 
likely be met with continued concern about the general level of 
federal spending, the largest block of expected funding for the 
California project is uncertain. 

State and Local Funding: 

The Authority is also relying on a total of about $8.2 billion in 
state high-speed rail bond proceeds (which includes the $3.7 billion 
that has been appropriated to date) and another $5 billion in locally-
generated funds for the project's construction.[Footnote 54] The 
proceeds from the high-speed rail bonds are dedicated, in that they 
can only be used for this project and do not have to compete with 
other budgetary priorities. However, the remaining $4.5 billion will 
have to be appropriated to the project. The $5 billion in local funds--
most of which are expected at the end of the project's construction 
timeline--have not been committed by local entities yet. Authority 
officials told us that these local funds could include revenues 
derived from property development in and around high-speed rail 
stations or improved service on existing transit corridors. For 
example, planned improvements on the Caltrain corridor may result in 
increased ridership; if so, a portion of increased revenues could be 
earmarked for high-speed rail construction. 

Cap-and-Trade Funding: 

The Authority acknowledges the risk in relying on a significant amount 
of federal funding (and to a lesser extent, local funding) for the 
project. To help mitigate this risk, the California Department of 
Finance has identified California's newly implemented cap-and-trade 
program as a potential source of revenue in the event other public 
funding is not available.[Footnote 55] However, there are a number of 
challenges with using this program as a potential source of revenue 
for the high-speed rail project. First, while the state of California 
anticipates this program to generate tens of billions of dollars in 
revenues, amounts are not certain since the auctions have only 
recently begun. The first auction of greenhouse gas emissions 
allowances was held in November 2012 and resulted in revenues at the 
low-end of the state's initial estimates.[Footnote 56] The state 
anticipates revenue will increase significantly by fiscal year 2015 
when more bidders are required to participate. Second, in order for 
any program revenue to benefit the high-speed rail project, 
appropriations must be enacted to provide funding for the project. 
This means that high-speed rail will have to compete with other state 
funding priorities, and the Authority cannot be assured what portion 
of cap-and-trade revenues, if any, will be dedicated for high-speed 
rail. Finally, concerns have been raised about whether revenues from 
the cap-and-trade program can be used for the high-speed rail project. 
An analysis by California's Legislative Analyst's Office (LAO) 
indicated that, among other things, the project might not comply with 
the legislative goal of the cap-and-trade program of reducing 
greenhouse gas emissions by 2020, which raises concerns about the 
potential use of cap-and-trade revenues for the high-speed rail 
project. The LAO suggested that the California Legislature obtain 
legal advice to address any potential legal issues. However, according 
to the California Air Resources Board's 2008 plan to implement the cap-
and-trade program, high-speed rail was identified as an eligible 
project to meet the state's greenhouse gas reduction goals. 

Securing Private-Sector Investment Will Depend on Reliable Operating 
Cost and Ridership Forecasts: 

In addition to the challenges of obtaining public-sector funding, the 
Authority may face challenges in attracting private-sector funding if 
its' operating cost estimate and ridership forecasts prove to be 
optimistic. The Authority expects that once the initial operating 
segment is operational in 2023, it will generate a profit that would 
be attractive to private investors. The Authority is planning to raise 
approximately $13.1 billion by selling an operating concession to a 
private firm or consortium of firms.[Footnote 57] The Authority plans 
to use the proceeds of this sale to help complete construction of the 
system. 

Our past work on high-speed rail systems has shown that private sector 
investment is easier to attract only after the public sector has made 
a substantial capital investment in the system.[Footnote 58] The 
Authority's plan is consistent with this funding approach; however, to 
successfully attract private investment in this project, the Authority 
will have to meet two significant milestones: 

* complete construction of the IOS (this will require at least an 
additional $25 billion of investment from public sources), and: 

* demonstrate that the IOS can operate at a profit.[Footnote 59] 

Public-private partnerships for intercity passenger rail, such as what 
the Authority is planning for, have been proposed but not implemented 
in the United States.[Footnote 60] However, in other transportation 
sectors, public-sector infrastructure owners have used public-private 
partnerships to incorporate private-sector operating expertise and 
encourage private investment. For example, the state of Indiana raised 
$3.8 billion in 2006 by arranging a private operating concession for 
its existing Indiana Toll Road. In addition, according to DOT, the 
city of Denver, Colorado, raised nearly $100 million from private 
sources to help finance a $2 billion expansion to its transit and 
commuter rail project. While the private-sector can provide needed 
funding and management expertise to a transportation project, this 
approach is not risk-free. As we have previously reported for highway 
projects, public-private partnership can also present trade-offs, such 
as the risk that the private operator will demand more revenue from 
users (e.g., tolls) than initially expected.[Footnote 61] 

Other nations' experiences with high-speed rail indicate that under 
certain circumstances the private sector can operate these systems and 
that they could potentially be profitable on an operating basis. For 
example, Japan sold certain high-speed rail lines to private operators 
and does not provide operating subsidies to these firms. And, in 2010, 
Britain sold an operating concession for its High-Speed 1 line to a 
consortium of private investors for approximately Ł2 billion 
(approximately $3.2 billion).[Footnote 62] Private firms have also 
expressed interest in operating high-speed rail projects in the United 
States. For example, according to FRA, several private consortia were 
preparing to submit bids on a HSIPR-funded project between Tampa and 
Orlando, Florida. However, public-sector support was withdrawn when 
the governor canceled the project which precluded private-sector 
investment. And, in Texas, a privately-financed high-speed rail 
project failed when the investors encountered financial difficulties. 
Authority officials told us that they have met with a number of 
private firms and high-speed rail operators that have expressed 
interest in California's project but have not entered into any 
agreements since the project has not yet been built. Attracting 
private investment may not only require up-front public investment but 
may also require the use of revenue guarantees, or public guarantees 
of a minimal level of income to the project regardless of ridership 
levels. Such guarantees will reduce the risk for private operators, 
and therefore their cost of raising capital, but according to the 
Authority, a revenue guarantee is considered to be a type of operating 
subsidy that is barred by the legislation that authorized the state 
high-speed rail bonds (Proposition 1A bonds). 

Private-sector investment in the California high-speed rail project, 
if any, will ultimately be determined by the profitability of the 
system--that is, the extent to which operating revenues exceed 
operating costs. The Authority currently estimates an operating profit 
in the first and all subsequent years of operation.[Footnote 63] 
However, this estimate is only as reliable as the underlying operating 
cost and revenue forecasts. As discussed earlier, the Authority's 
current ridership and revenue forecasts are reasonable for planning 
purposes, however, further refinements will be required as the project 
continues to evolve. The Authority's current operating cost estimates 
will also need to be improved in the future. Accordingly, both cost 
and ridership forecasts will change before the initial operating 
segment is completed in 2022, making the future value of potential 
private funding uncertain at this time. 

The Authority Comprehensively Identified Potential Economic Impacts, 
but It Is Too Early to Determine Specific Impacts: 

Potential Economic Impacts Were Comprehensively Identified, but 
Specific Impacts Will Depend on Project Decisions, Economic 
Conditions, and Other Factors: 

The Authority comprehensively identified key economic impacts that 
could result from the high-speed rail project, including user and non-
user impacts, as required by FRA and other federal requirements. FRA 
guidance, as contained in the program Notices of Funding Availability 
(NOFA), requires HSIPR applicants to identify the potential benefits 
and costs of proposed projects with a focus on a public return on 
investment.[Footnote 64] A public return on investment includes a 
project's potential to deliver transportation, economic recovery, and 
other benefits. To assist in project evaluation, FRA encouraged HSIPR 
applicants to provide an economic analysis that quantified the 
monetary value of user benefits and, if available, public benefits. 
[Footnote 65] However, according to FRA, program NOFAs did not 
explicitly require either a formal benefit-cost analysis (BCA) or 
preparation of an economic impact analysis (EIA).[Footnote 66] FRA 
officials said their review of economic impacts was based on a 
reasonableness test--that is, were economic impacts identified and 
were the assumptions behind the impacts reasonable. The officials said 
that FRA did not have the time or resources to conduct an in-depth 
analysis and that a reasonableness test provided increased assurance 
as to potential economic impacts of project proposals. Projects 
awarded funding must also comply with the National Environmental 
Policy Act (NEPA) and its implementing regulations. NEPA requires that 
government agencies undertaking a major federal action (such as 
providing grant funding) with significant effects on the environment 
prepare an analysis of the environmental impacts of the proposed 
action, including a discussion of alternatives to the proposed 
action.[Footnote 67] Under FRA's guidelines for considering 
environmental impacts, among the impacts to be considered in a NEPA 
environmental assessment are such things as land use and potential 
economic effects on existing business districts and metropolitan 
areas. The program NOFAs required HSIPR applicants seeking funds to 
develop new high-speed rail corridors and intercity passenger rail 
services to complete a NEPA review and the June 2009 program NOFA 
required HSIPR applicants to present information that provided a 
business and investment justification that contained project cost and 
benefit estimates. 

To analyze the extent to which economic impacts were comprehensively 
identified by the Authority, the key components of economic impacts on 
users and non-users were identified from a variety of sources, 
including NEPA, NOFAs for the HSIPR and Transportation Investment 
Generating Economic Recovery (TIGER) programs as well as a report on 
best practices for assessments of public benefits prepared for the DOT 
OIG for the DOT OIG's March 2012 report on FRA's HSIPR program. 
[Footnote 68] The best practices report identified user impacts such 
as travel costs and travel quality and non-user impacts such as 
highway and airway congestion and noise or air quality levels. Non-
user impacts may also include economic impacts that are difficult to 
quantify, such as the effects improved transportation connectivity may 
have on allowing firms to access larger labor or product markets or 
increasing the labor supply because people can more easily access 
jobs. The latter are included in what is termed "wider economic 
impacts." FRA's requirement for a public return on investment from 
HSIPR projects included aspects of both user and non-user impacts. The 
best practices report noted the importance of both ridership and 
revenue forecasts and cost estimates in determining public benefits. 
In particular, it stated that public benefits assessment depends 
heavily on ridership and revenue forecasts and the implications these 
have on project impacts on travelers and the general population. 
Similarly, operating, maintenance, and capital cost estimates were 
also identified as important elements of public benefits assessments. 

The Authority's April 2012 revised business plan identified the 
primary user and non-user economic impacts of the California high-
speed rail project (see table 5). For example, the plan identified 
potential user impacts such as travel time reliability for high-speed 
rail users and non-user impacts such as the effects on highway 
congestion and economic development around stations. In addition to 
the business plan, the Authority prepared a BCA that provided a more 
detailed analysis of both user and non-user impacts included in the 
revised business plan. The Authority also prepared an EIA that focused 
on those other economic impacts of the system that do not fall into 
the BCA framework. According to the Authority, the EIA presented 
longer-term impacts on California's economy from building the high-
speed rail system. 

Table 5: Primary Economic Impacts Associated with High-Speed Rail 
Identified by the Authority in the April 2012 Revised Business Plan, 
BCA, or EIA: 

User impacts: 

Primary economic impacts associated with high-speed rail: Generalized 
travel costs to users. Includes impacts on the following: 
* In-vehicle travel time; 
* Waiting, access, egress time; 
* Out-of-pocket cost; 
Economic impacts identified by the Authority: 
* In-vehicle travel time; 
* Access and egress times; 
* Vehicle operating costs (fuel and non-fuel such as operating and 
maintenance costs); 
Where economic impact discussed: Revised Business Plan; BCA. 

Primary economic impacts associated with high-speed rail: Travel 
quality. Includes crowding, comfort, travel time variance and 
reliability of service; 
Economic impacts identified by the Authority: 
* Comfort of high-speed rail; 
* Reliability of travel times; 
Where economic impact discussed: Revised Business Plan; BCA. 

Non-user impacts: 

Primary economic impacts associated with high-speed rail: Congestion. 
Includes reduced congestion on the following: 
* Highway; 
* Air; 
* Rail; 
Economic impacts identified by the Authority: 
* Congestion impacts on highway vehicle miles traveled; 
* Congestion impacts on aviation (flight and passenger delays); 
* Analysis excluded potential impacts in urban rail corridors near Los 
Angeles and the San Francisco Bay Area; 
Where economic impact discussed: Revised Business Plan; BCA. 

Primary economic impacts associated with high-speed rail: Environment. 
Includes impacts on the following: 
* Noise; 
* Air quality; 
* Carbon emissions; 
* Landscape/townscape; 
Economic impacts identified by the Authority: 
* Automobile and truck emissions impacts; 
* Automobile noise pollution; 
* Aviation emissions impacts; 
* BCA analysis specifically excluded impacts on land use and land 
values; 
Where economic impact discussed: BCA. 

Primary economic impacts associated with high-speed rail: Travel 
safety; 
Economic impacts identified by the Authority: 
* Automobile accident cost impacts; 
Where economic impact discussed: BCA. 

Primary economic impacts associated with high-speed rail: Financial 
impacts, including impacts on such things as: 
* Operator revenues; 
* Federal, state, and local tax revenues; 
Economic impacts identified by the Authority: 
* Potential impacts on state gross domestic product from investing in 
high-speed rail; 
* Potential impacts on local tax revenue; 
Where economic impact discussed: Revised Business Plan; EIA. 

Source: GAO analysis of Steer Davies Gleave and Authority documents. 

[End of table] 

Although the Authority comprehensively identified the primary 
potential economic impacts, it is too early to determine specific 
economic impacts since these will depend on a number of factors. These 
include the following: 

* Future project decisions. The California high-speed rail project is 
in its early stages of development, and a number of project decisions 
have yet to be made, including final alignment of train routes, some 
station locations, and the type and frequency of service. Decisions 
such as these can be expected to have a bearing on potential economic 
impacts. For example, route alignments and station locations can 
affect economic development. While acknowledging that the extent to 
which high-speed rail would change California's economic landscape was 
not fully understood, the EIA suggested that based on studies in other 
countries, the main economic impacts from high-speed rail in 
California will likely occur in areas within 2 hours of major economic 
centers, such as the San Francisco Bay Area and Los Angeles. However, 
the EIA also concluded that the greatest volume of redevelopment 
attributable to high-speed rail will likely occur in major 
metropolitan areas and that the Central Valley could see moderate 
clustering of development around stations. Although economic 
development around stations offers the potential for economic impacts, 
achieving such development may be subject to a number of factors, and 
certain impacts may not be easy to identify. In July 2012, we reported 
on the potential for economic development associated with bus rapid 
transit projects.[Footnote 69] We found that in the five case study 
locations we examined, although the bus rapid-transit project was 
having some positive effect on economic development, individuals 
associated with these projects were unsure about how much economic 
activity could be attributed to the presence of bus rapid transit 
compared with other factors or circumstances. In addition, the project 
sponsors and experts we spoke with told us that transit-supportive 
policies and development incentives can play a crucial role in helping 
to attract and spur economic development associated with bus rapid 
transit. In September 2009 we reported the characteristics of transit-
oriented development around stations can increase nearby land and 
housing values, but we also found that determining transit-oriented 
effects on the availability of affordable housing in these 
developments are complicated by lack of direct research and data. 
[Footnote 70] 

* Economic conditions over the life of the project. As we reported in 
March 2012, the closer the economy is to full employment, the smaller 
net effect a project will have on total economic activity.[Footnote 
71] The speed with which the nation and California recover from the 
2007 to 2009 recession, which cannot be known in advance, will affect 
the net employment from any new infrastructure projects. If the 
economy achieves full employment, such projects would affect the 
composition of employment but not its level or rate of growth. 
[Footnote 72] The high-speed rail project will be constructed and 
operated over a period of many decades and likely over many economic 
cycles. The Authority's April 2012 BCA used a 67-year period (from 
2013 to 2080) to estimate potential economic benefits and costs of the 
project, and, for purposes of analyzing potential operating and 
maintenance costs, the April 2012 revised business plan assumed a 38-
year operating period (2022 to 2060).[Footnote 73] Over such an 
extended period, economic conditions can be expected to change, as 
will potential economic impacts. The Authority's April 2012 EIA 
recognizes that the project's economic impacts will be affected by 
California's economy and unemployment rates. According to the 
Authority's EIA, as of February 2012 California's 10.9 percent 
unemployment rate was the nation's third highest.[Footnote 74] It goes 
on to estimate that the high-speed rail project has the potential to 
create about 1 million direct and indirect job-years through 
construction of the Phase 1 blended system, based on the assumption 
that 20,000 job-years would be created for each $1 billion in capital 
investment.[Footnote 75] The accuracy of this estimate will depend on 
the economic conditions and unemployment rates at the time the jobs 
are created. The EIA acknowledged this uncertainty when it stated that 
multipliers used to estimate indirect and induced jobs are snapshots 
in time of an economy and represent only current or recent economic 
relationships and technologies. They do not capture structural changes 
in the economy, new technologies, or changes in wages that occurred 
since the multiplier data were produced or that might occur in the 
future. 

* Uncertainty about some impacts. Uncertainty may particularly affect 
non-user impacts, some of which are difficult to measure or quantify. 
The Authority's April 2012 EIA discusses the potential for wider 
economic impacts from high-speed rail such as the benefits of bringing 
California's economic activities and markets closer together by 
reducing travel times. The EIA states, by improving transportation 
connectivity and reducing congestion, the high-speed rail system could 
make California's economy more efficient, productive, and competitive 
by such things as bringing businesses closer to labor and other 
markets and providing workers with greater access to jobs. However, 
the best practices report prepared for the DOT OIG found that 
additional data on things like the relationship between economic 
density and productivity, labor supply elasticity, and price-cost 
margins were needed to assess wider economic impacts of high speed 
rail projects.[Footnote 76] The study went on to note that no research 
was currently available regarding wider economic impacts of U.S. high-
speed rail projects since such projects do not yet exist. The 
Authority's April 2012 EIA recognized this difficulty and states that 
the extent to which the high-speed rail project will affect the 
economic landscape of California is not well understood, though 
transportation infrastructure investments have historically created 
fundamental shifts in the spatial relationship between places. These 
limitations do not imply deficiencies in Authority or DOT performance 
but rather the inherent analytical complexities of large 
infrastructure investments. 

* Uncertainty regarding local or regional impacts. The April 2012 EIA 
discusses the potential for economic impacts along the high-speed rail 
line, including direct and indirect employment opportunities, 
increased efficiencies and productivity from bringing labor and other 
markets closer together, and transit-oriented development around 
stations. The specific impacts to regions or localities will depend on 
a number of factors, including project-related factors and factors 
associated with local policies and decisions. Among the project-
related factors are both the rate of project spending over time as 
well as where project funds are spent. The high-speed rail project is 
expected to be constructed over a long period of time and in phases 
when funding becomes available. The rate of spending and its timing 
will influence when and to what extent regions and localities may 
experience economic impacts associated with the project. Which 
specific regions or localities may experience economic impacts will be 
influenced by such project decisions as route alignments and station 
locations. 

Local policies and decisions will also affect regional and local 
economic impacts. Studies conducted for the Authority by the 
University of California at Berkeley suggested there are opportunities 
for economic development from the high-speed rail system in a variety 
of locations, including Fresno and Bakersfield in California's Central 
Valley. However, the studies cautioned that the extent of economic 
development will depend on cities establishing a framework of planning 
and development policies that encourage development. Some cities have 
begun taking actions to promote economic development related to the 
high-speed rail system. For example, in June 2012 Fresno issued a 
solicitation for consultants to prepare a master plan for its planned 
downtown rail station that would enable the city to maximize local 
economic benefits from the high-speed rail system. Other cities have 
not yet acted. The Authority's April 2012 EIA found that of the 13 
potential stations on the Phase I blended corridor, 7 (53.8 percent) 
did not have station-specific development plans. 

We also found some limitations in the specific economic analyses 
prepared by the Authority. In particular, the April 2012 BCA has 
shortcomings that could limit its usefulness to decision makers. 

* Identification of negative impacts of the project and their effects 
on the BCA analysis. The BCA lacked detail regarding the handling of 
negative impacts associated with the high-speed rail project. DOT 
guidance on benefit-cost analysis suggests that negative (or adverse) 
impacts, such as non-user (or highway) delays associated with rail 
construction, be included in BCA analyses to facilitate consistent 
project comparisons.[Footnote 77] The guidance recommends that 
negative impacts, such as highway delays associated with rail 
construction, be shown as a negative benefit and not included in 
project investment costs to better facilitate comparisons between 
projects. The April 2012 BCA recognized that during the period of 
project construction there would be roadway delays in urban areas that 
would offset some travel time savings. However, the BCA excluded such 
impacts from the analysis since the impacts were expected to be (1) 
localized, (2) minimal since the high-speed rail project minimizes 
urban grade crossings, and (3) negligible in proportion to overall 
travel time savings once the project is complete. 

Aside from roadway delays, the Authority acknowledged there were 
additional categories of economic impacts that may be negative. For 
example, the April 2012 revised business plan discusses how the high-
speed rail system could limit access to parts of farmland in the 
Central Valley potentially reducing the output of affected farmlands. 
The BCA contained little discussion of such impacts and states that 
the BCA did not incorporate or monetize land use and land value 
impacts the high-speed rail project may cause (positive or negative). 
The Authority said negative impacts were assumed to be part of the 
mitigation measures that would be conducted as part of the 
environmental review process and right-of-way acquisition and that the 
costs of these measures were included in the cost side of the benefit-
cost calculation. According to the Authority, including negative 
impacts as a negative benefit would lead to double counting them as 
both a negative benefit as well as a cost in the benefit-cost 
calculation. We agree that negative impacts of the high-speed rail 
project should not be double counted, but the BCA should include 
discussion of potential negative impacts and how they are treated in 
the analysis. Such information would better inform decision makers 
about the existence of negative impacts and their potential effect on 
project benefits or costs. 

* Identification of the risks and uncertainties associated with the 
BCA analysis. The BCA did not discuss the potential risks and 
uncertainties associated with either the benefits or costs used in the 
analysis. Forecasts are inherently uncertain, including those for 
ridership and economic projections. Recognition and analysis of risks 
is an important part of project evaluation. As we reported in February 
2011, Executive Order 12893 and Office of Management and Budget (OMB) 
Circulars Nos. A-94 and A-4 indicate that benefit and cost information 
shall be used in agency decision making and that the level of 
uncertainty in estimates of benefits and costs shall be disclosed. 
[Footnote 78] In particular, Executive Order 12893 requires that 
uncertainties about the amount and timing of important benefits and 
costs associated with an infrastructure investment be recognized and 
addressed through appropriate quantitative and qualitative assessments. 

Similarly, DOT's TIGER guidance, which the Authority used to prepare 
its BCA, requires applicants to assess the reliability of any 
forecasts used to generate benefit estimates but does not specifically 
require a discussion of risks and uncertainties in a BCA. We reported 
in February 2011 that the majority of the applications to the TIGER 
and HSIPR programs we reviewed did not provide information related to 
uncertainties in projections, data limitations, or the assumptions 
underlying their models.[Footnote 79] Even when such information was 
provided, we found that it was not always comprehensive. We 
recommended that DOT require, among other things, grant applicants to 
clearly communicate the level of uncertainty in estimates of project 
benefits and costs. The Authority did not conduct a risk analysis 
beyond examining the potential effects of high and low cost scenarios. 
While the Authority may have used credible sources for variables in 
its analysis (e.g., fuel prices), this does not eliminate all 
forecasting risks, and those risks should be identified. In addition, 
as we have noted, decisions on route alignments and other aspects of 
the project have yet to be made; decisions that could add to project 
risk. 

High-Speed Rail's Impact on Meeting State Travel Demand Has Not Been 
Fully Assessed: 

Although the Authority comprehensively identified the project's 
potential economic impacts, additional analysis is needed at the state 
level of how high-speed rail will affect other transportation modes 
and their ability to meet future travel demand. This includes the 
potential cost of additional improvements that may be needed or 
conversely planned projects that may not be needed. An important 
aspect of non-user impacts is a project's potential effect on other 
transportation modes, including highways, aviation, and local transit 
systems. This is an important issue since the Authority has estimated 
that, as a result of population growth and other factors, overall 
interregional trips in California will increase from about 500 million 
in 2000 to about 900 million in 2030. In addition, under the blended 
approach adopted by the Authority, success of the high-speed rail 
system will depend in part on local transit improvements. 

As part of the planning process, the Authority considered high-speed 
rail's impact on the capacity of other transport modes. For example, 
in April 2012 the Authority issued an analysis of the highway and 
airport improvements that would be needed to provide an equivalent 
capacity to the high-speed rail system envisioned in the April 2012 
revised business plan.[Footnote 80] The analysis found the total cost 
(in 2011 dollars) of equivalent capacity investment in highways and 
airports would range between $123 billion and $138 billion to build up 
to 4,600 highway lane-miles, 115 airport gates, and 4 airport runways. 
[Footnote 81] However, the analysis did not focus on potential 
additional highway or other transportation improvements that may be 
required even with construction of the high-speed rail system. Rather, 
the analysis identified the potential highway and airport improvements 
that would be required to provide an equivalent capacity to that of 
the high-speed rail, not an assessment of additional improvements 
required to meet future intercity travel demand. Identifying such 
improvements was not the task of the Authority since its task is to 
develop a high-speed rail system. Rather, this task would fall to the 
state as part of its overall planning responsibilities under federal 
transportation-planning requirements. 

Constructing a high-speed rail system is not expected to meet all of 
California's future intercity travel demand. Among other things, 
development of the phase I blended approach will affect the need for 
additional improvements to local transportation systems to support 
high-speed rail. For example, officials with the Orange County 
Transportation Authority told us they have bus projects and a street 
car project in the planning phases that will be linked to both local 
commuter rail and the high-speed rail system. Similarly, officials 
from the Southern California Association of Governments told us the 
cost of integrating high-speed rail with local transit was still being 
developed and that the cost of an initial list of prioritized projects 
to facilitate this integration exceeded $3 billion. Of this $3 billion 
in projects, about $1 billion in projects were categorized as high 
priority. The officials said that although they have got agreement 
from the Authority to help fund some of the high priority projects and 
some funding is expected to come from Proposition 1A, a funding source 
for the list of potential projects had not been determined. 

Even though high-speed rail is not expected to meet all of 
California's future intercity travel demand, statewide transportation 
planning has not yet fully assessed the impact of the high-speed rail 
system in meeting this demand. In November 2011, the California 
Transportation Commission (CTC) issued a Statewide Transportation 
System Needs Assessment[Footnote 82] that identified the preservation, 
management, and expansion projects required over the 2011 to 2020 
period. The assessment identified a total cost for the projects of 
about $540 billion and a nearly $300 billion funding gap in meeting 
the project needs identified. According to CTC officials, the needs 
assessment included information about high-speed rail that was readily 
available from Authority documents. However, high-speed rail was not 
assessed in terms of project needs, costs, or the funding gap because, 
according to CTC officials, Regional Transportation Plans that formed 
the basis of the assessment did not include high-speed rail in 
identifying project needs and costs. Similarly, California Department 
of Transportation officials told us current highway transportation 
planning has not looked at this issue, and the department's long range 
transportation plans have not included consideration of high-speed 
rail. According to the officials, the department did not see an 
immediate need to do an assessment since the high-speed rail system is 
not expected to be operational for another 10 years or more. The 
officials agreed that the high-speed rail system will affect highways, 
and its impact will need to be considered in future transportation 
plans. 

Conclusion: 

As currently conceived, California's high-speed rail project is 
expected to be among the most expensive infrastructure projects that 
has been undertaken in this country. Therefore, concerns about 
potential cost escalations and optimistic ridership forecasts, as well 
as the potential burden this could place on public budgets are well 
placed. Cost and revenue estimates for large projects are, by their 
nature, imprecise; these estimates endeavor to predict many years into 
the future within the confines of what is known today. According to 
our past work reviewing high-speed rail projects in other countries, 
cost and ridership estimates tend to be overly optimistic. However, 
experts agree that taking steps to anticipate project risks and 
improve the credibility of such estimates will lower the risk of cost 
overruns and missed revenue forecasts. Improving the reliability of 
cost and revenue forecasts is critical to providing project sponsors, 
FRA, the Congress, and ultimately, the public with greater confidence 
that this project can be viable. This confidence is of particular 
importance as the Authority will seek significant and sustained 
funding from federal, state, and private sources. 

We found that the Authority did not fully employ best practices for 
producing reliable cost estimates as expressed in GAO's Cost Guide, 
which are recommended practices but not required. The cost estimates 
can be improved as the project progresses from design to construction, 
and ultimately, to operation. The Authority was not required to follow 
the Cost Guide; instead, it was required to follow FRA's guidance 
which we found to be limited. That guidance identified the cost 
categories that applicants should include in its cost analyses, but 
did not specify how cost estimates should be generated. The Authority 
told us that it looked to FTA's cost-estimating guidance to help 
inform the Authority's cost-estimating methodology. The Authority can 
be commended for supplementing its analyses using FTA's guidance, but 
this does not necessarily ensure a fully reliable cost estimate. Our 
past work, as well as that of the DOT OIG, has shown that FRA has yet 
to develop sufficient program guidance for project evaluation and 
oversight under its additional grant-making responsibilities under the 
HSIPR program. Developing guidance for HSIPR applicants and grantees 
that incorporates best practices from the Cost Guide would allow cost 
estimators to improve the reliability of cost estimates for expensive 
projects like high-speed rail. Such guidance would help ensure project 
costs reflect the four characteristics required for developing 
reliable cost estimates and minimize the risk of cost overruns, missed 
deadlines, and unmet performance targets. 

The Authority is in the process of updating its ridership and revenue 
model in response to recommendations provided by experts and peer 
review groups, such as the Ridership and Revenue Peer Review Panel. We 
believe that these steps will have the potential to improve the 
Authority's ridership forecasts, and we encourage continued 
refinement, as more information becomes available and continued review 
by peer review groups. Improved forecasts will be particularly 
important as the Authority seeks to secure private investment in the 
project. The potential project revenues--which are primarily dependent 
on ridership--will help determine how much the Authority may be able 
to obtain from private sources. Similar to the cost estimates, FRA has 
developed minimal guidance for applicants to develop reliable 
ridership and revenue forecasts. We, along with the DOT OIG, have 
previously made recommendations that FRA improve this guidance to 
ensure reliability of ridership and revenue forecasts that are used to 
determine the viability of high-speed rail projects. According to FRA 
officials, the agency is currently in the process of implementing 
these recommendations. Since we have already made recommendations to 
FRA on this issue in our prior work and FRA is taking actions on these 
recommendations, we are not at this time making additional 
recommendations related to improving FRA's ridership and revenue 
forecasting guidance. 

Recommendation: 

We recommend that the Secretary of Transportation direct the 
Administrator of FRA to improve its guidance for high-speed rail 
project sponsors to better ensure that cost estimates that are 
submitted by applicants seeking federal funding are accurate, 
comprehensive, well-documented, and credible according to the best 
practices detailed in GAO's Cost Guide. 

Agency Comments and Our Evaluation: 

We provided a draft of this report to DOT and the Authority for review 
and comment. DOT neither agreed nor disagreed with our recommendation. 
In an e-mailed response, DOT said it was pleased that the Authority 
met many of the criteria in the Cost Guide for producing accurate, 
comprehensive, well-documented, and credible cost estimates; that 
ridership and revenue forecasts were reasonable; and that the 
Authority did a comprehensive job in identifying potential economic 
impacts. However, DOT noted (1) that the currently funded project has 
sound cost estimates while future, currently unfunded phases will 
continue to be refined as the project progresses and data improves, 
(2) that FRA's cost-estimating guidance was the best available at the 
time, and (3) that GAO's Cost Guide focuses on federally managed 
acquisitions. DOT's response noted that the project is a multi-decade 
effort consisting of many segments and phases, each in a different 
stage of development. DOT's response also noted that the Cost Guide 
was issued in March 2009, one month after passage of the Recovery Act 
and a few months before deadlines for the HSIPR guidance, and, 
therefore, it was not feasible for FRA to incorporate Cost Guide best 
practices into guidance. Finally, DOT said applying the Cost Guide 
principles to future FRA capital cost guidance, while feasible, would 
require analysis and adaptation to accommodate unique aspects of long-
term grantee-managed transportation projects. DOT noted that the Cost 
Guide is focused primarily on federally managed acquisitions and 
programs, not infrastructure projects that non-federal parties will 
develop and build. DOT also provided technical comments that we 
incorporated as appropriate. 

We recognize DOT' concerns; however, our charge was to assess the 
reliability of the cost estimates and not just whether they complied 
with FRA's guidance, which we found to be less than best practices. 
While the Cost Guide was released in 2009, it is a culmination of cost-
estimating best practices that have previously been published and have 
been available to federal agencies for many years. Therefore, these 
practices could have been considered when preparing HSIPR program 
guidance. Finally, the best practices contained in the guide are 
applicable to developing cost estimates for a wide variety of programs 
and projects, whether federally managed or not. 

The Authority provided a letter summarizing its comments about the 
report (see appendix IV). In general, the Authority believes that the 
report highlights its efforts to produce cost estimates that reflect 
the scope of the project, that methods and models used to develop 
ridership and revenue forecasts adhered to applicable best practices, 
and that comprehensively identifying potential economic impacts 
demonstrates a strong economic case for the project. However, the 
Authority noted that different components of its program (such as 
implementation phases and construction packages) are at different 
stages of development and that it would not be practicable to apply 
the full complement of tools in the Cost Guide at the program level at 
this time. This is because the environmental review is under way on a 
number of sections and alignments and other choices are still to be 
made. The Authority also stated it plans to improve its cost estimates 
and ridership forecasts. For example, the Authority stated that the 
updated ridership model being developed for the 2014 business plan 
will incorporate many of the changes we suggested and that the 
Authority will improve the quantification of project risks. 

We commend the Authority for planning to improve its cost estimates 
and forecasts. Regarding not applying the Cost Guide at the program 
level, we note that a program does not need to be in an advanced stage 
of planning in order to complete a sensitivity or cost risk and 
uncertainty analysis. In fact, such analyses are most valuable when 
performed early in a program's life cycle. Single point estimates are 
more uncertain at the beginning of a program because less is known 
about its detailed requirements and the opportunity for change is 
greater. For example, undefined or unknown technical information, 
uncertain economic conditions, and political issues are often 
encountered during a program's acquisition. For management to make 
good decisions, the program estimate must reflect the degree of 
uncertainty, so that a level of confidence can be given about the 
estimate. Therefore, it is important to conduct a risk and uncertainty 
analysis at all stages of a project so cost estimates reflect the risk 
and uncertainty that exist. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution of this report 
until 30 days from the report date. At that time, we will send copies 
of this report to the appropriate congressional committees, the 
Secretary of Transportation, the Administrator of FRA, and the 
Director of OMB. The report will also be available at no charge on the 
GAO website at [hyperlink, http://www.gao.gov]. 

If you or your staff have any questions about this report, please 
contact me at (202) 512-2834 or flemings@gao.gov. Contact points for 
our Offices of Congressional Relations and Public Affairs may be found 
on the last page of this report. GAO staff who made major 
contributions to this report are listed in appendix V. 

Signed by: 

Susan A. Fleming: 
Director: 
Physical Infrastructure Team: 

List of Addressees: 

The Honorable Bill Shuster: 
Chairman: 
The Honorable Nick Rahall: 
Ranking Member: 
Committee on Transportation and Infrastructure: 
House of Representatives: 

The Honorable Jeff Denham: 
Chairman: 
The Honorable Corrine Brown: 
Ranking Member: 
Subcommittee on Railroads, Pipelines and Hazardous Materials: 
Committee on Transportation and Infrastructure: 
House of Representatives: 

The Honorable Darrell Issa: 
Chairman: 
House Committee on Oversight and Government Reform: 
House of Representatives: 

The Honorable Karen Bass: 
House of Representatives: 

The Honorable John Campbell: 
House of Representatives: 

The Honorable Jim Costa: 
House of Representatives: 

The Honorable Mike Honda: 
House of Representatives: 

The Honorable Duncan Hunter: 
House of Representatives: 

The Honorable Barbara Lee: 
House of Representatives: 

The Honorable Zoe Lofgren: 
House of Representatives: 

The Honorable Kevin McCarthy: 
House of Representatives: 

The Honorable Tom McClintock: 
House of Representatives: 

The Honorable Howard P. "Buck" McKeon: 
House of Representatives: 

The Honorable John Mica: 
House of Representatives: 

The Honorable Gary Miller: 
House of Representatives: 

The Honorable Devin Nunes: 
House of Representatives: 

The Honorable Adam Schiff: 
House of Representatives: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

This report assesses (1) the reliability of the California High-Speed 
Rail Authority's (Authority) estimates of the project's costs; (2) the 
reasonableness of the Authority's passenger rail ridership and revenue 
forecasts; (3) the risks to the Authority's plan to fund the project; 
and (4) the comprehensiveness with which the Authority identified 
potential economic impacts of the project. Our analysis focused on the 
Authority's cost estimates, ridership and revenue forecasts, and 
economic estimates presented in the April 2012 revised business plan. 

To address these objectives, we reviewed numerous documents, including 
Federal Railroad Administration (FRA) guidance, Department of 
Transportation Office of Inspector General (DOT OIG) reports, prior 
GAO reports, and pertinent legislation. In addition, we obtained 
documents from and conducted interviews with Authority officials to 
obtain information about the Authority's process for developing its 
various cost estimates, ridership and revenue forecasts, and economic 
impact estimates. Specifically, we reviewed the November 2011 draft 
business plan and April 2012 revised business plans and documentation 
used to develop the analysis presented in those plans. We conducted 
interviews with the Authority's contractors--Parsons Brinckerhoff and 
Cambridge Systematics--to obtain additional information about the 
Authority's processes for developing these estimates and to clarify 
information in their written documentation. In addition, we also 
conducted interviews with officials from various federal and state 
agencies, peer review groups, academic experts, advocacy groups, and 
transit and local government groups to obtain information on, among 
other things, their role with the California high-speed rail project 
and their views on the Authority's cost estimates, financing plans, 
ridership and revenue forecasts, and potential economic impacts. (See 
table 6 for a list of organizations and individuals we interviewed for 
this study.) 

Table 6: Organizations and Individuals Interviewed: 

Federal agencies: 
Department of Transportation, Federal Aviation Administration;
Department of Transportation, Federal Highway Administration;
Department of Transportation, Federal Railroad Administration;
Department of Transportation, Federal Transit Administration;
Department of Transportation, Office of Inspector General. 

State agencies:
California Air Resources Board.
California Bureau of State Audits;
California High-Speed Rail Authority;
California Legislative Analyst's Office;
California State Treasurer;
California Transportation Commission;
Caltrans. 

Peer review groups: 
California High-Speed Rail Peer Review Group;
Ridership and Revenue Peer Review Panel. 

Advocacy groups and experts:
Californians for High-Speed Rail;
Citizens Advocating Responsible Rail Design;
Samer Madanat, Mark Hansen, and David Brownstone, University of 
California Berkeley Institute of Transportation Studies;
Arpad Horvath and Mikhail Chester, University of California Berkeley 
Civil and Environmental Engineering;
Wendell Cox;
Julien Dehornoy;
Bent Flyvbjerg;
Joseph Vranich. 

Transit and local government groups: 
Amtrak;
Bay Area Rapid Transit;
Caltrain;
Los Angeles County Metropolitan Transportation Authority;
Metropolitan Transportation Commission;
Orange County Transportation Authority;
Peninsula Corridor Joint Powers Board;
San Francisco County Transportation Authority;
Southern California Association of Governments;
Southern California Regional Railroad Authority;
Transbay Transit Center Joint Powers Authority. 

Source: GAO. 

[End of table] 

To assess the reliability of the project cost estimates, we analyzed 
the Authority's cost estimating approach against GAO's best practices 
found in the 2009 GAO Cost Estimating and Assessment Guide (Cost 
Guide).[Footnote 83] GAO designed the Cost Guide to be used by federal 
agencies to assist them in developing reliable cost estimates and also 
as an evaluation tool for existing cost estimates. To develop the Cost 
Guide, GAO cost experts assessed measures applied by cost-estimating 
organizations throughout the federal government and industry and 
considered best-practices for the development of reliable cost-
estimates. We analyzed the cost estimating practices used by the 
Authority against these best practices. For our reporting needs, we 
collapsed these best practices into four general categories 
representing practices that help ensure that a cost estimate is (1) 
accurate, (2) well documented, (3) comprehensive, and (4) credible. 

After a review of all source data, including but not limited to 
electronic cost models for both capital investment and operating and 
maintenance phases, all supporting documentation, personal interviews, 
and independent research, we assessed the extent to which the 
Authority met these best practices on a five-point scale: 

* Not Met--Authority provided no evidence that satisfies any of the 
criteria. 

* Minimally Met--Authority provided evidence that satisfies a small 
portion of the criterion. 

* Partially Met--Authority provided evidence that satisfies about half 
of the criterion. 

* Substantially Met--Authority provided evidence that satisfies a 
large portion of the criterion. 

* Fully Met--Authority provided complete evidence that satisfies the 
entire criterion. 

We determined the overall assessment rating by assigning each 
individual rating a number: Not Met = 1; Minimally Met = 2; Partially 
Met = 3; Substantially Met = 4; and Fully Met = 5. For the purposes of 
this assessment we have also included a Not Applicable (N/A) 
assessment category. Then, we took the average of the individual 
assessment ratings to determine the overall rating for each of the 
four characteristics. The resulting average becomes the Overall 
Assessment as follows: Not Met = 0 to 1.4; Minimally Met = 1.5 to 2.4; 
Partially Met = 2.5 to 3.4; Substantially Met = 3.5 to 4.4; and Fully 
Met = 4.5 to 5.0. 

To assess the reasonableness of the Authority's ridership and revenue 
forecasts, we analyzed the extent to which the Authority's methods for 
developing the ridership model and resulting ridership and revenue 
forecast adhered to federal guidance and generally accepted travel 
demand modeling practices for high-speed rail projects. Unlike with 
GAO's cost-estimating criteria discussed earlier, there is no single 
industry standard for developing or evaluating intercity passenger 
high-speed rail ridership forecasts. As such, for the purposes of our 
assessment, we reviewed a variety of sources that identify generally 
accepted travel demand modeling practices and developed criteria based 
on these practices to assess the reasonableness of the approach used 
to create the ridership and revenue models for the California high-
speed rail project.[Footnote 84] In developing our criteria, we relied 
primarily on a 2011 report prepared for the DOT OIG's office by the 
firm Steer Davies Gleave, on best practices related to developing high-
speed rail ridership and revenue forecasts.[Footnote 85] The report 
provides a description of current standard practices in high-speed 
rail ridership and revenue forecasting, key steps typically involved 
in completing these forecasts, and a description on the range of data 
and methods used in the forecasting process. The intent of this 
guidance is to provide information that will assist reviewers to 
understand and evaluate forecasting studies. In addition, we also 
examined other literature on developing rail ridership and revenue 
forecasts to corroborate information in the Steer Davies and Gleave 
report. Specifically, we reviewed, among other sources, forecasting 
guidance from the FRA, Federal Highway Administration (FHWA), prior 
GAO reports and other ridership and revenue guidance in academic 
research. (See table 7 for a list of sources used to develop criteria). 

Table 7: Guidance and Reports Used to Identify Generally Accepted 
Travel-Demand-Modeling Practices: 

Federal agency guidance: 

Source: Prepared by Steer Davies Gleave for the Department of 
Transportation, Office of Inspector General; 
Guidance or report: HSIPR Best Practices: Ridership and Revenue 
Forecasting (April 2011). 

Source: Prepared by Cambridge Systematics for the Federal Highway 
Administration; 
Guidance or report: Travel Model Validation and Reasonableness 
Checking Manual, Second edition (September 2010). 

Source: Department of Transportation, Federal Railroad Administration; 
Guidance or report: High-Speed Intercity Passenger Rail (HSIPR) 
Program; Notice of Funding Availability; issuance of interim program 
guidance,"74 Federal Register 119" (23 June 2009), pp. 29900-29929. 

Guidance or report: High-Speed Intercity Passenger Rail Program, 
Notice of Funding Availability of Individual Projects (NOFA); 
issuance of interim program guidance, "75 Federal Register 126" (1 
July 2010), 38365-38383. 

Prior GAO reports: 

Guidance or report: Intercity Passenger and Freight Rail: Better Data 
and Communication of Uncertainties Can Help Decision Makers Understand 
Benefits and Trade-offs of Programs and Policies, [hyperlink, 
http://www.gao.gov/products/GAO-11-290]; (Washington D.C.: Feb. 24, 
2011). 

Guidance or report: High Speed Passenger Rail: Future Development Will 
Depend on Addressing Financial and Other Challenges and Establishing a 
Clear Federal Role, [hyperlink, 
http://www.gao.gov/products/GAO-09-317]; (Washington D.C.: Mar. 19, 
2009). 

Academic Literature: 

Guidance or report: Bent Flyvbjerg, "Quality Control and Due Diligence 
in Project Management: Getting Decisions Right by Taking the Outside 
View," International Journal of Project Management (November 2012). 

Guidance or report: Transportation Research Board, "Metropolitan 
Travel Forecasting: Current Practice and Future Direction, Special 
Report 288", (2007). 

Guidance or report: McFadden, Bolduc, Bierlaire, "The Estimation of 
Generalized Extreme Value Models from Choice-Based Samples", School of 
Architecture, Civil and Environmental Engineering, Ecole Polytechnique 
Federale de Lausanne (August, 2006). 

Guidance or report: The British Department for Transport, "Procedures 
for Dealing with Optimism Bias in Transport Planning", (June 2004). 

Guidance or report: Flyvberg, Bruzelius, Rothengatter, "Megaprojects 
and Risk: An Anatomy of Ambition," Cambridge University Press (2003); 
SourceFlyvberg, Bruzelius, Rothengatter, "Megaprojects and Risk: An 
Anatomy of Ambition," Cambridge University Press (2003). 

Source: GAO. 

[End of table] 

From these sources we identified recommended practices related to 
seven key steps relating to: (1) developing trip tables, (2) 
determining and applying service characteristics, (3) developing mode 
choice models, (4) estimating induced travel, (5) estimating expected 
fare revenue, (6) conducting sensitivity analysis, and (7) conducting 
validation testing. We compared generally accepted practices for each 
of these steps to the Authority's process for developing the ridership 
and revenue forecast as outlined in the April 2012 revised business 
plan and in supporting technical documentation. We could not evaluate 
each of the many detailed design decisions, assumptions, and model 
inputs used by the Authority, but rather focused on the seven key 
steps and whether they were implemented in accordance with generally 
accepted practices. We reviewed documents from and conducted 
interviews with Authority officials and their contractor--Cambridge 
Systematics--to obtain information about the Authority's process for 
developing the ridership and revenue forecasts. Specifically, we 
examined the Authority's process for developing the models used to 
produce the various forecasts, the assumptions and data sources used 
to develop the models, the survey instruments used to collect data, 
and the Authority's process for model estimation, calibration, and 
validation. 

We focused our analysis on identifying key steps in developing 
ridership forecast models for high-speed rail projects, elements 
affecting validity and reliability of models, common limitations of 
models and pitfalls, and recommended approaches for external review. 
In addition, we interviewed organizations that had conducted reviews 
of the Authority's ridership and revenue forecasts, such as academic 
experts from University of California Berkeley's Institute of 
Transportation Studies (ITS) and members of the Ridership and Revenue 
Peer Review Panel and the California High-Speed Rail Peer Review 
Group. From these interviews, we obtained additional information about 
(1) generally accepted methods used for project ridership and revenue 
for high-speed projects and elements of these approaches that have the 
greatest potential risk, (2) general assumptions underlying demand 
forecast models, elements impacting validity and reliability of 
models, and existing data limitations, and (3) the extent to which the 
Authority's approach follows generally accepted practices for 
developing valid and reliable ridership and revenue estimates. 

To assess the Authority's financing plan, we reviewed the plan, 
conducted interviews with Authority officials and other state and 
federal officials, and reviewed literature and other information on 
financing for high-speed rail projects in other countries as well as 
large transportation projects in the United States. 

To assess how well the Authority identified economic impacts 
associated with the high-speed rail project, we reviewed the April 
2012 revised business plan as well as the April 2012 benefit-cost 
analysis and April 2012 economic impact analysis. In addition, to 
establish criteria for the various components of economic impact 
analysis, including user and non-user impacts, we reviewed pertinent 
legislation, such as the Passenger Rail Investment and Improvement Act 
of 2008 and the National Environmental Policy Act, and the NOFAs 
associated with the HSIPR and Transportation Investment Generating 
Economic Recovery (TIGER) grant programs. FRA officials told us the 
NOFA's outlined the type of information HSIPR grant applicants were to 
provide regarding project benefits and costs and how this information 
would be reviewed by FRA in reviewing grant applications. We also 
reviewed reports from the DOT OIG regarding HSIPR project viability 
assessments. In particular, we reviewed the June 2011 report prepared 
for the DOT OIG's office on best practices related to public benefit 
assessments of high-speed rail projects produced by the firm Steer 
Davies Gleave. The Steer Davies Gleave report identified important 
components of user and non-user impacts associated with public 
benefits assessments. 

To gain a better understanding of economic impact analysis, we 
reviewed the Economic Impact Analysis Primer prepared by the FHWA's 
Office of Asset Management. The Economic Analysis Primer identified 
the basic process of identifying and analyzing economic impacts, 
including benefit-cost analyses. The document also identified the 
similarities and differences between benefit-cost analyses and 
economic impact analyses. We also interviewed officials from FRA, the 
Authority, FHWA, the Federal Transit Administration, DOT OIG, and 
Steer Davies Gleave about economic impact issues. To assess issues 
related to high-speed rail and future travel demand, we reviewed the 
Authority's revised business plan and April 2012 equivalent capacity 
study as well as the November 2011 Statewide Transportation Systems 
Needs Assessment prepared by the California Transportation Commission 
(CTC). We also discussed these issues with officials from CTC and the 
California Department of Transportation as well as officials with 
local transportation agencies in California about potential 
improvement projects they had planned that were associated with the 
high-speed rail project. 

The proposed high-speed rail project is a very large public works 
project with costs expected to be spread over more than a decade. 
Depending on how cost figures are presented, different impressions of 
the magnitude and funding requirements of the program could be given. 
Whether or not the effects of inflation are included in the estimate 
is a source of significant differences. Year of expenditure (YOE) 
dollars include inflation in out-year costs, a convention adopted to 
facilitate budgeting over time but not necessarily a good 
representation of the true economic costs of the project. Removing the 
increase in cost attributable solely to inflation in the price level 
provides a better picture of burden on taxpayers and other funders 
because the tax base, including incomes, property values, and retail 
sales, would have increased with inflation as well. In the case of the 
high-speed rail project, the YOE cost total is 25 percent greater than 
that when inflation effects are removed. An estimate of the cost in 
present value terms, which accounts for inflation and the time value 
of money, would be smaller still. 

We conducted this performance audit from February 2012 to March 2013 
in accordance with generally accepted government auditing standards. 
These standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

[End of section] 

Appendix II: GAO Analysis of California High-Speed Rail Authority's 
Cost Estimates: 

Characteristic: Accurate; 
Overall assessment[A]: Substantially Met; 

Best practice: The cost estimate results are unbiased, not overly 
conservative or optimistic, and based on an assessment of most likely 
costs; 
Individual assessment: Partially Met; 
While the Authority has attempted to ensure accuracy and eliminate 
bias in their estimate by conducting sensitivity analysis, parametric 
checks, and the use of peer review, these have all been on subsets of 
the total program. No risk or sensitivity analysis has been developed 
at the program level or between the low and high estimates. 
Alternative high and low estimates do not create a range of estimates, 
but rather point estimates evolving around potential options. In the 
absence of cost risk and uncertainty analysis it is not possible to 
determine if the estimate is unbiased. Unless the estimate is based on 
an assessment of the most likely costs and reflects the degree of 
uncertainty given all of the risks considered, management will not be 
able to make informed decisions. 

Best practice: The estimate has been adjusted properly for inflation; 
Individual assessment: Substantially Met; 
Both capital investment and operations and maintenance (O&M) costs are 
inflated to YOE dollars using sound data and methodologies. Source 
data used for cost estimating are normalized to appropriate base 
years, although in some instances the normalizing processes were not 
clear. 

Best practice: The estimate contains few, if any, minor mistakes; 
Individual assessment: Fully Met. 

Best practice: The cost estimate is regularly updated to reflect 
significant changes in the program so that it is always reflecting 
current status; 
Individual assessment: Fully Met. 

Best practice: Variances between planned and actual costs are 
documented, explained, and reviewed; 
Individual assessment: Not applicable. 

Best practice: The estimate is based on a historical record of cost 
estimating and actual experiences from other comparable programs; 
Individual assessment: Substantially Met; 
The estimate relies on construction cost data from commercial 
databases heavily supplemented with local construction bids from 
analogous construction projects. The Authority collects technical and 
summary-level cost data on existing and future high-speed trainsets, 
but there is no documentation that explains how these data were 
adjusted for use in the cost estimate. The O&M estimate relies on 
applicable historical data. However, the extent of applicability is 
unknown because adjustments are not thoroughly documented. 

Best practice: The estimating technique for each cost element was used 
appropriately; 
Individual assessment: Substantially Met; 
The estimating techniques are reasonable for those Standard Cost 
Categories (SCC) elements discretely estimated, where a unit price 
estimating methodology was employed. The 2012 O&M model is a 
simplified version of the 2009 model, appropriately suited to the cost 
model's stated purpose of establishing program viability. However, the 
simplification results in an unnecessary loss of fidelity in some cost 
elements. 

Characteristic: Comprehensive; 
Overall assessment[A]: Partially Met; 

Best practice: The cost estimate includes all life cycle costs; 
Individual assessment: Substantially Met; 
The Authority has included all relevant costs with the relatively 
minor exclusion of disposition costs in the capital investment 
estimate. 

Best practice: The cost estimate completely defines the program, 
reflects the current schedule, and is technically reasonable; 
Individual assessment: Partially Met; 
The technical baseline description for the capital investment cost 
estimate resides in multiple documents that collectively comprise the 
technical baseline of the program. However, there is no distinct 
technical baseline description for the O&M estimate. Officials stated 
that later versions of the O&M estimate will align with the Concept of 
Operations plan, which was approved in February 2012. 

Best practice: The cost estimate work breakdown structure (WBS) is 
product-oriented, traceable to the statement of work/objective, and at 
an appropriate level of detail to ensure that cost elements are 
neither omitted nor double-counted; 
Individual assessment: Partially Met; 
The program utilized the FRA (SCC) and associated definitions for the 
capital investment costs. The cost estimate expands upon this 
structure to provide detailed identification of infrastructure work, 
but has reduced insight into common support costs. The standardized 
O&M FRA SCC elements were not used for capturing O&M costs because the 
O&M estimate was not required to comply with the SCC elements. While 
the O&M estimate includes common elements for administration and 
support costs, the O&M WBS is greatly simplified. As a consequence, up 
to two-thirds of O&M costs are collected in a single cost element. 

Best practice: The estimate documents all cost-influencing ground 
rules and assumptions; 
Individual assessment: Partially Met; 
Ground rules and assumptions are imbedded in much of the documentation 
for both the capital investment and O&M estimates as well as in the 
cost models, but not all assumptions have supporting rationale or 
sources. As the design for a specific section advances, risks are 
quantified and assigned to specific WBS elements. At the program 
level, contingency factors are used to capture less-defined risks. 
However, O&M risks are not specifically related to O&M WBS elements, 
and the impact of budget constraints on specific WBS elements has not 
been clearly defined. In addition, the impacts of technology maturity 
on cost are not entirely defendable. Unless ground rules and 
assumptions are clearly documented, the cost estimate will not have a 
basis for areas of potential risk to be resolved. 

Characteristic: Well documented; 
Overall assessment[A]: Partially Met; 

Best practice: The documentation should capture the source data used, 
the reliability of the data, and how the data were normalized; 
Individual assessment: Partially Met; 
The documentation provides some insight into the development of the 
cost estimates; however, much of our analysis was based on information 
derived from interviews and discussions with Authority 
representatives, not from readily available information in the 
documentation. The O&M model includes relevant data, but sources and 
variables can only be described as somewhat documented. For the most 
part, documentation relates how inputs are adjusted from past O&M 
models but fails to account for how earlier values were derived. 
Without sufficient background knowledge about the source and 
reliability of the data, the cost estimator cannot know with any 
confidence whether the data collected can be used directly or need to 
be modified. 

Best practice: The documentation describes in sufficient detail the 
calculations performed and the estimating methodology used to derive 
each element's cost; 
Individual assessment: Substantially Met; 
The documentation provided varying degrees of insight into the 
estimating methodology. The majority of costs--that is, infrastructure 
and site work--are described at a detailed level by unit cost, 
quantities, labor rates, equipment, and material costs, and the like. 
However, some cost elements had little or no supporting documentation. 

Best practice: The documentation describes step-by-step how the 
estimate was developed so that a cost analyst unfamiliar with the 
program could understand what was done and replicate it; 
Individual assessment: Partially Met; 
Details of the estimating process and methodology were provided for 
the track structure and track and site work elements of the capital 
investment model, but supporting data and details of how other 
elements are estimated were not available. No comprehensive document 
exists that explains the O&M model element by element. Without good 
documentation, management and oversight will not be convinced that the 
estimate is credible. In addition, analysts unfamiliar with the 
program will not be able to replicate the estimate because they will 
not understand the logic behind it. 

Best practice: The documentation discusses the technical baseline 
description and that the data in the baseline are consistent with the 
estimate; 
Individual assessment: Partially Met; 
The documentation of the capital investment cost model and the 
technical baseline are consistent with one another. The primary 
emphasis and underlying data sources are for the infrastructure and 
site work, but little definition or supporting data are provided for 
the remaining cost elements. In addition, the O&M cost estimate is not 
based on an approved technical baseline document, although officials 
state that later versions will be aligned to the Concept of Operations 
plan. Because the technical baseline is intended to serve as the basis 
for developing a cost estimate, it should be discussed in the cost 
estimate documentation. 

Best practice: The documentation provides evidence that the cost 
estimate was reviewed and accepted by management; 
Individual assessment: Partially Met; 
Documents indicate that the Authority's management team was engaged in 
reviewing the cost estimates and there are multiple indications that 
management reviewed pieces of the cost estimate. However, many of 
these reviews appear to be for subsets of the total program, either by 
construction package or phase, and focusing more on the financing 
rather than the detailed estimating methodology or underlying 
assumptions. While specific subsets of the estimate appeared to be 
reviewed by or discussed with management, we found no specific 
instance where the total program estimate, including supporting source 
data and estimating methodologies, was provided to senior management 
for review, discussion, and subsequent approval; 
Because a cost estimate should form the basis for establishing the 
budget, it is imperative that management understands how the estimate 
was developed, including the risks associated with source data and 
estimating methodologies. 

Characteristic: Credible; 
Overall assessment[A]: Partially Met; 

Best practice: The cost estimate includes a sensitivity analysis that 
identifies a range of possible costs based on varying major 
assumptions, parameters, and data inputs; 
Individual assessment: Partially Met; 
A formal sensitivity analysis has been performed for Contract Package 
1, design and construction of the first 26-33 miles of trackway 
infrastructure between the counties of Madera and Fresno. In addition, 
the Authority conducted limited sensitivity analysis on summary-level 
variables in the O&M model. However, sensitivity analysis of the 
entire program estimate has not been done. The capital investment 
estimate includes low and high cost alternative alignments, and the 
O&M estimate provides three alternative scenarios driven by ridership 
options. However, without a complete sensitivity analysis that reveals 
how the cost estimate is affected by a change in a single assumption, 
the cost estimator will not fully understand which variable most 
affects the cost estimate. 

Best practice: A risk and uncertainty analysis was conducted that 
quantified the imperfectly understood risks and identified the effects 
of changing key cost driver assumptions and factors; 
Individual assessment: Partially Met; 
The Authority utilized FRA guidance in developing its estimates, 
guidance that does not require risk or uncertainty analysis at the 
program level this early in the design stage. Authority officials 
stated that more advanced engineering designs are being developed to 
support the process and that risk and uncertainty analysis has been 
undertaken on Contract Package 1. Authority officials acknowledge the 
existence of risk and have tried to accommodate expected risk through 
the application of contingency factors. While the capital investment 
and O&M models include a contingency element, the factors used do not 
appear to be based on historical data or analogous sources; 
Lacking risk and uncertainty analysis, management cannot determine a 
defensible level of contingency reserves that are necessary to cover 
increased costs resulting from unexpected design complexity, 
incomplete requirements, technology uncertainty, and other 
uncertainties. 

Best practice: Major cost elements were crossed checked to see whether 
results were similar; 
Individual assessment: Partially Met; 
The Authority recognizes the importance of crosschecks and identified 
a series of crosschecks to verify and validate the results of the 
data. Authority officials stated there are several stages of 
crosschecks and quality control which are described in their cost 
estimating procedures. Yet little documentation has been provided that 
would allow us to verify that crosschecks and alternative 
methodologies have been developed. For example, estimators have 
crosschecked major cost factors in the O&M model with cost data from 
foreign systems, but there is no evidence that costs have been 
estimated using different methodologies. 
The main purpose of cross-checking is to determine whether alternative 
methods produce similar results. If so, then confidence in the 
estimate increases, leading to greater credibility. The Authority has 
contracted with the International Union of Railways (UIC) for a study 
that intends to verify and validate the capital investment cost model. 
Authority officials stated that the UIC panel of experts will also 
provide a set of international cost comparisons for infrastructure 
maintenance and rolling stock maintenance. 

Best practice: An independent cost estimate was conducted by a group 
outside the acquiring organization to determine whether other 
estimating methods produce similar results; 
Individual assessment: Minimally Met; 
An independent cost estimate (ICE) was performed on the Merced-Fresno 
and Fresno-Bakersfield segments for infrastructure costs. However, 
while the segments cover 35 percent of the planned system rail length, 
they make up less than 10 percent of the overall estimated program 
cost. An ICE should be performed on the entire program, including O&M 
costs. ICEs can provide decision makers with additional insights into 
a program's potential costs because they frequently use different 
methods and are less burdened with organizational bias. 

Source: GAO. 

[A] Not Met--The Authority provided no evidence that satisfies any of 
the criterion, Minimally Met--The Authority provided evidence that 
satisfies a small portion of the criterion, Partially Met--The 
Authority provided evidence that satisfies about half of the 
criterion, Substantially Met--The Authority provided evidence that 
satisfies a large portion of the criterion, and Fully Met--The 
Authority provided complete evidence that satisfies the entire 
criterion. 

[End of table] 

[End of section] 

Appendix III: Description of Generally Accepted Travel-Demand-Modeling 
Practices and Authority's Methods for Developing Ridership and Revenue 
Forecasts: 

Travel-demand-modeling tasks: Developing trip tables; 
Generally accepted travel-demand-modeling practices: A central task of 
the ridership forecasting process involves collecting and compiling 
data on current travel patterns into trip tables. This process 
involves the following: 
* Collection of trip table data: In order to calculate the number of 
trips that divert to the new or improved high-speed rail mode, the 
level of trip making without the new or improved high-speed rail mode 
must first be established. Reliable information on modal travel 
volumes is a prerequisite for valid ridership and revenue forecasts. 
The competing modes from which high-speed rail draws its share are 
mainly auto, rail, bus and air. There are various methods for 
collecting trip table data for these modes, including, obtaining 
publicly available data collected through travel surveys as well as 
from special-purpose surveys developed and undertaken for the purpose 
of a particular transportation study. Data can also be obtained from 
trip distribution models and ticket sales; 
* Base and forecast year trip table development: The base and forecast 
year input trip tables are the basis for a study's ridership estimates 
and revenue forecasts. Any overestimate or underestimate of the trip 
tables will translate to high or low forecasts of ridership. Base trip 
tables generally summarize the current total number of trips by mode 
for each city pair along the route and are generally prepared by using 
a variety of sources of data on actual trip making patterns. Growth 
factors--which determine the rate of increase over time--can then be 
applied to the base trip tables to develop forecast year trip tables, 
which contain estimates of future travel on various modes in the 
absence of a proposed high-speed rail alternative. Forecast year trip 
tables may also be prepared by estimating future-year trips directly; 
* Trip segmentation: Trip tables are generally segmented by mode of 
travel, trip purpose, and other traveler characteristics. Criteria 
frequently used in defining market segments include trip purpose, trip 
length, traveler income, travel party size, and others; 
Authority's methods for developing ridership and revenue forecasts: 
* The Authority collected data from a variety of sources including, 
among others, socioeconomic data from local agencies, U.S. Census 
Bureau, and the California Department of Finance; travel data from 
various travel surveys; and highway, air, conventional rail, and urban 
transit network data from local agencies; 
* The high-speed ridership and revenue model for inter-regional travel 
was developed utilizing surveys and other statewide travel 
information. Intra-regional travel models from Metropolitan Planning 
Organizations (MPOs) in the San Francisco and Los Angeles regions were 
adapted for use in the high-speed rail ridership and revenue model 
from the models maintained by the MPOs for those regions. A factoring 
process was used to estimate ridership in the San Diego region; 
* Base year trip tables were developed from existing California 
regional models used by local authorities including the Metropolitan 
Transportation Commission (MTC) and the Southern California 
Association of Governments as well as interregional trip tables 
developed from travel survey data. Forecast year trip tables were 
developed by projecting base year forecast data to forecast year 2030,; 
* and then the models were run on 2030 projections; 
* Trip were segmented by long versus short trips (over and under 100 
miles), and trip purposes (commute, business, recreation, other). 

Travel-demand-modeling tasks: Determining and applying service 
characteristics; 
Generally accepted travel-demand-modeling practices: Travel demand 
forecasts require information on the level of service (LOS) 
characteristics such as travel time and fares, of competing modes of 
travel, such as travel time for automobile and air travel, along the 
proposed route and of the high-speed rail alternative. LOS data are 
required for all modes in the study (i.e., air, rail, auto) and for 
base and forecast years. This information is used in the mode choice 
modeling process, whereby service characteristics of the various 
available modes are compared to the service characteristics of the 
proposed high-speed rail mode. This allows forecasters to estimate the 
likelihood that travelers will divert from their existing travel mode 
to high-speed rail; 
Determining and applying service characteristics for a proposed 
network involves: 
* Developing a network representation: Network representation can be 
detailed (i.e., with detailed representations of street and transit 
networks that include location, alignment, connections, and service 
characteristics) or can be less explicit and instead focus directly on 
zone to zone level of service data. A less explicit network 
representation can be used if the structure of the network is very 
simple); 
* Preparing skim tables: Skim tables contain data on the time, cost, 
and other service characteristics of the various modes that are 
available for a trip; 
Accurate and realistic representation of the base and forecast year 
LOS characteristics is of paramount importance for realistic high-
speed ridership forecasting. Rail LOS information may be approximately 
derived from the service plan but may not represent it in complete 
detail; 
Authority's methods for developing ridership and revenue forecasts: 
* The Authority developed a detailed network representation for the 
entire state to forecast travel between regions. Data were obtained 
from the existing statewide highway network and details were added 
using data from local regional models, from the MTC, the Southern 
California Association of Governments, the San Diego Association of 
Governments, and data from the Kern County region; 
* LOS characteristics were defined for the four inter-regional travel 
modes: auto, conventional rail, high-speed rail, and air. LOS 
characteristics covered three broad categories: costs, times, and 
reliability, which were summarized in travel skim tables. Several of 
these characteristics were varied during model application to see how 
ridership and revenue would be impacted. Characteristics were 
collected from published or observed data from various sources 
including, the MTC and the Federal Aviation Administration. The high-
speed rail characteristics were based on the initial service plan and 
fare structure. 

Travel-demand-modeling tasks: Developing a mode choice model; 
Generally accepted travel-demand-modeling practices: Mode choice 
modeling predicts the outcome of the decision process by which 
travelers choose the mode(s) to take from origin to destination. This 
step predicts the fraction of trips that will divert to the high-speed 
rail project from existing modes. Mode choice models are typically 
developed via a statistical analysis of the behavior of travelers in 
different situations. The statistical analysis may be based on 
behavior observed in actual travel situations (revealed preference 
data) or in hypothetical situations presented to travelers in a survey 
(stated preference data), or both; 
Selection of the mode choice model structure is important to obtain 
reliable and credible ridership and revenue estimates. Two types of 
mode choice models are commonly used in high-speed rail demand 
forecasting: 
* Choice models: These models are used to predict the decisions of 
travelers considering alternative transportation modes. Multinomial 
logit models and nested logit model are types of choice model that can 
be used; 
* Diversion choice model: A diversion choice model considers only two 
modes-the one in use in the base situation and the high-speed rail 
alternative; 
Authority's methods for developing ridership and revenue forecasts: 
The Authority developed two choice models: intra-regional urban model 
(models behavior associated with shorter distance and more frequent 
trip making) and an inter-regional model (models traveler behavior 
associated with longer-distance travel); 
* Intra-regional (Urban) Models: For both the San Francisco Bay Area 
and the greater Los Angeles regions, mode choice models were adapted 
from existing models to include the high-speed rail mode. The updated 
mode choice models were applied using the MPO trip tables for each 
region as input. San Diego is the only other region that contains the 
possibility of intra-regional high-speed rail trips, but the estimate 
of these riders was very low relative to the other regions. Because 
the level of effort to develop, calibrate, and apply the regional mode 
choice model was very high, intra-regional ridership for San Diego was 
developed using a population-based estimate rather than a traditional 
mode choice model; 
* Inter-regional models: The Authority developed four sets of models 
which included trip frequency, destination choice, primary mode 
choice, and access/egress mode choice. The destination choice 
component predicts the destinations of the trips generated in the trip 
frequency component based on zonal characteristics and travel 
impedances. The mode choice components (main mode choice, access mode 
choice and egress mode choice) predict the modes that the travelers 
would choose based upon the modal service levels as well as 
characteristics of the travelers and trips being made; 
* Data were derived from, among other sources, the California 
Department of Transportation Statewide Model, existing regional mode 
choice models, and revealed preference and stated preference survey 
data. The economic and household characteristics were forecast for 
each zone in the year 2030 based on data and forecasts from state, 
regional, and local government agencies. The primary main mode choice 
model relied primarily on stated preference data. 

Travel-demand-modeling tasks: Estimating induced travel; 
Generally accepted travel-demand-modeling practices: A high-speed rail 
project will improve the overall level of service for intercity travel 
within a given corridor. This improvement will make conditions more 
favorable for travel. Trips will therefore be taken on high-speed rail 
that might not otherwise have been made using any of the current 
modes. The new trips are commonly referred to as induced travel; 
An upper limit on induced travel of approximately 10 percent of total 
high-speed rail trips is widely accepted for proposed high-speed rail 
systems in the U.S; 
Authority's methods for developing ridership and revenue forecasts: 
The Authority forecasted 2.05 percent induced travel for the blended 
Phase 1 low scenario. 

Travel-demand-modeling tasks: Estimate expected fare revenue; 
Generally accepted travel-demand-modeling practices: Expected fare 
revenue is determined by a calculation using the ridership estimates 
generated by the model and the average fares. The total ridership for 
the system is generally calculated by adding the diverted trips 
calculated from the mode choice models and the induced trips to 
produce the total ridership for the system; 
Authority's methods for developing ridership and revenue forecasts: 
The Authority calculated fare revenue by multiplying the ridership 
estimates generated from the ridership model by the average high-speed 
rail fares forecasted for each region-to-region pair. 

Travel-demand-modeling tasks: Conduct sensitivity testing; 
Generally accepted travel-demand-modeling practices: All ridership and 
revenue forecasting studies should incorporate an analysis of the 
sensitivity of forecast results to key inputs and modeling assumptions 
including fare, running time, service frequency, station locations, 
and assumptions about socio-economic and travel growth in forecast 
years. Sensitivity analysis typically is conducted by varying, more or 
less systematically, selected forecasting model inputs, parameters, or 
assumptions (e.g., inflation rate or fuel cost) around their 
"standard" value, running the model, and examining the variation in 
outputs. Sensitivity analysis can help determine the reliability 
associated with the model output forecasts and can help identify the 
factors that have greatest impact on project ridership and revenue; 
Authority's methods for developing ridership and revenue forecasts: 
Several sensitivity tests were done to determine how the model reacted 
to different sets of assumptions such as changes to fuel costs, travel 
times, and fares. In addition, the Authority developed an extreme case 
scenario to test the sensitivity of the model to a series of downside 
events, such as increased average rail travel time from Merced to the 
San Fernando Valley (140 min. instead of 126 min), decreased train 
frequency (3 trains per hour instead of 4 trains per hour during peak 
times), lower auto-operating costs and lower air fares (10 percent 
below actual 2009 average air fares). 

Travel-demand-modeling tasks: Conduct model validation; 
Generally accepted travel-demand-modeling practices: Model validation 
is a key component of ridership and revenue forecasting and generally 
consists of testing the validity of the model using data other than 
(and usually newer than) the data from which it was estimated, to 
assess how well the model predicts actual ridership; 
There are two superior (but not often performed) ways of checking 
model performance: (1) the historical method, in which a prior-year 
model is used to forecast current travel, which is then compared with 
actual current travel; and (2) "backcasting," in which a current year 
model is used to estimate travel for a prior year, which is then 
compared with actual travel in the prior year. Backcasting is used by 
5 percent of all and 13 percent of large MPOs; 
Authority's methods for developing ridership and revenue forecasts: 
The Authority validated the model through tests performed using 
Amtrak's Acela service in the Northeast Corridor (NEC) as input to the 
California high-speed rail model and compared the output with 2008 
actual ridership and 2030 NEC forecasts. Efforts to validate the model 
by comparing to the NEC appear reasonable. The NEC is not an ideal 
test for the model but it is the only one available in the U.S. Use of 
foreign systems would raise difficult issues of comparability. 

Source: GAO, DOT OIG, FHWA, FTA, and academic studies of ridership and 
revenue best practices and Authority information. 

[End of table] 

[End of section] 

Appendix IV: Comments from the California High-Speed Rail Authority: 

California High-Speed Rail Authority: 
770 L Street, Suite 800: 
Sacramento, CA 95814: 
916-324-1541: 
[hyperlink, http://www.cahighspeedrail.cs.gov] 

March 8, 2013: 

Board Members:
Dan Richard, Chairperson;
Lynn Schenk Vice-Chairperson;
Thomas Richards, Vice Chairperson;
Jim Hartnett;
Michael Rossi;
Thomas J. Umberg;
Jeff Morales, Chief Executive Officer. 

Ms. Susan Fleming:
Director, Physical Infrastructure:
U.S. Government Accountability Office: 
441 G Street, NW: 
Washington D.C. 20548: 

Re: Response to Government Accountability Office Draft Report 13-304, 
"California High-Speed Passenger Rail: Project Estimates Could be 
Improved to Better Inform Future Decisions" (GAO Code 541090). 

Dear Ms. Fleming: 

The California High-Speed Rail Authority (Authority) appreciates the 
Government Accountability Office's (GAO) professionalism and 
thoroughness in conducting the most extensive external review to date 
of the program's cost estimates and other analyses. This letter 
highlights some of the findings that we found particularly important 
and discusses the areas where we have begun to take steps to 
incorporate your findings into our on-going efforts to improve our 
estimates. 

As the GAO states in the Draft Report, the Authority was not required 
to follow the GAO Cost Guide and did follow all applicable guidance 
from the Federal Railroad Administration (FRA), which is administering 
the Authority's federal grants. Where no guidance from the FRA 
existed, the Authority followed Federal Transit Administration (FTA) 
guidance. As guidance from the FRA and FTA continues to evolve, we 
will continue to meet all applicable requirements. 

As noted by the GAO, different components of the Authority's program 
(such as the implementation phases and construction packages) are at 
different stages of development and analysis. Given this, it would not 
be practicable to conduct every analysis in the Cost Guide at the 
program level at this time. That is because the environmental review 
is underway on a number of sections with alignment and other choices 
still to be made. As the program advances, the individual "projects" 
that comprise the overall program will all undergo each of the steps 
described in the Cost Guide. 

The GAO's finding that "the Authority substantially met best practices 
for developing accurate cost estimates," highlights the Authority's 
efforts to produce cost estimates that correctly reflect the program's 
scope. Additionally, the Draft Report states that the "GAO found the 
Authority's ridership and revenue forecasts to be reasonable" and that 
the methods and model used to develop them adhered to applicable best 
practices. This finding reinforces conclusions from the Independent 
Peer Review Group and Ridership Peer Review Panel that the ridership 
forecasts are reliable. 

Further, the Draft Report found that "The Authority comprehensively 
identified key economic impacts that could result from the high speed 
rail project," demonstrating the strong economic case for the project, 
including its ability to create jobs in the short-term and its 
potential to change the state's economic landscape in the long-term 
through improved connections between its metropolitan areas. 

While these findings are encouraging validation of the work that we 
have done to date, we fully acknowledge that any estimate can be 
refined and improved. As part of our effort to continue improving our 
estimates, we have already begun to incorporate many of the 
suggestions included in the GAO report. One such area is in the 
continued improvement of our ridership estimates. The updated 
ridership model being developed in time for the 2014 Business Plan 
will include many of the changes suggested by the GAO. The model will 
incorporate updated socioeconomic inputs and new data from survey work 
that is being conducted right now. These changes will continue to fine-
tune and refine the estimates based on the most up-to-date information 
available on characteristics of the travel market in California. 

At the same time, we continue to improve the quantification of the 
risks associated with each project that makes up the overall program. 
As alignments are selected and individual pieces of the program 
continue to advance through various stages of design and development, 
we will be able to conduct many of the risk and uncertainty analyses 
that are recommended in the GAO's Cost Guide. The risk quantification 
analysis will be integral to the procurement and business model 
decisions that will form the framework for private investment in the 
program. 

Finally, we are also working to upgrade our Operations and Maintenance 
(O&M) cost forecasting. For the 2014 Business Plan, we are developing 
an improved O&M cost model that includes a much more detailed bottom-
up cost estimate. That update is addressing issues raised in the GAO 
review, as well as detailed review and analysis by international high 
speed rail experts convened through the International Union of 
Railways, or MC. This updated model will be able to help validate the 
top-down estimate from the 2012 Business Plan and to test the 
financial impacts of different operating scenarios going forward. 

The process undertaken by the GAO over the last year has helped 
identify many of the areas where our estimates are strongest as well 
as areas of focus for continued improvement in the coming months and 
years. We will continue to incorporate GAO's input as we work to 
employ best practices in all aspects of the planning and delivery of 
our program. 

I thank you and your staff for your hard work and collaborative 
approach and we look forward to the release of the Final Report later 
this month. Please feel free to contact me with any questions at 
(916)384-1487. 

Sincerely, 

Signed by: 

Jeff Morales: 
Chief Executive Officer: 

[End of section] 

Appendix V: GAO Contact and Staff Acknowledgment: 

GAO Contact: 

Susan A. Fleming, (202) 512-2834 or flemings@gao.gov: 

Staff Acknowledgments: 

In addition to the individual named above, Paul Aussendorf, Assistant 
Director; Russell Burnett; Jason Lee; Delwen Jones; Richard Jorgenson; 
James Manzo (Technomics, Inc.); Maria Mercado; Susan Offutt; Paul 
Revesz; Max Sawicky; Maria Wallace; and Crystal Wesco made key 
contributions to this report. 

[End of section] 

Footnotes: 

[1] The National Railroad Passenger Corporation's (Amtrak) Acela 
service is currently the fastest passenger train service in the United 
States--it is capable of traveling at a maximum of 150 miles per hour 
but the average speed on the Washington, D.C., to Boston, 
Massachusetts, corridor is below 80 miles per-hour. 

[2] Estimates of the costs and revenues associated with the high-speed 
rail project appear in this report in two different formats: year of 
expenditure (YOE) and inflation adjusted dollars. YOE means the cost 
is presented in nominal dollars that reflect anticipated inflation 
over the period from the base year to the year in which a specific 
component of spending would take place. These estimates are also 
expressed in terms of current dollars in the base year of 2011, 
reversing the inflation adjustment used to calculate YOE dollars. The 
current dollar estimate will be lower than that expressed in YOE 
terms. Revenue estimates are expressed in current dollars in a base 
year. See appendix 1 for a discussion of the choice of methods for 
representing the costs of a project that spends out over a long 
period. Unless otherwise noted, all costs are presented in YOE dollars. 

[3] GAO, High-Speed Rail: Future Development Will Depend on Addressing 
Financial and Other Challenges and Establishing a Clear Federal Role, 
[hyperlink, http://www.gao.gov/products/GAO-09-317], (Washington, 
D.C.: Mar. 19, 2009). 

[4] Pub. L. No. 110-432, div. B, 122 Stat. 4848, 4960 (Oct. 16, 2008); 
Pub. L. No. 111-5, 123 Stat. 115, 208 (Feb 17, 2009). 

[5] Pub. L. No. 111-117, div. A, 123 Stat. 3034, 3056 (Dec. 16, 2009); 
Pub. L. No. 112-110, § 2222 (Apr. 15, 2011). 

[6] GAO Cost Estimating and Assessment Guide: Best Practices for 
Developing and Managing Capital Program Costs, [hyperlink, 
http://www.gao.gov/products/GAO-09-3SP], (Washington, D.C.: March 
2009). 

[7] Proposition 1A was codified in California law. See Cal. Sts. & 
High. Code § 2704 et seq. (2012). 

[8] This project will construct new rail right-of-way to provide 
service, some of which may require acquisition of privately owned land. 

[9] Program management services and ridership-and-revenue-modeling 
services for the California high-speed rail project are generally 
carried out by consultants under contract with the Authority. Parsons 
Brinckerhoff is the consulting firm responsible for program management 
of the California high-speed rail system. In addition, the Authority 
retained Cambridge Systematics--a transportation-consulting firm that 
provides ridership and modeling forecast services--to develop a 
ridership and revenue model for the project. 

[10] An additional $819.3 million was appropriated by the state 
legislature for connectivity projects and $252.6 million for 
environmental, system design, and preliminary engineering work. 

[11] In addition, $400 million was awarded to the Transbay Joint 
Powers Board for construction at the Transbay Transit Center in San 
Francisco. Pursuant to Proposition 1A, the Transbay Transit Center is 
the northern terminus of the California high-speed rail line. 

[12] The grant agreement between FRA and the Authority was amended in 
December 2012. The amendment revised the schedule of federal and state 
cash flows so that federal Recovery Act funds would be expended in 
advance of state funds. According to FRA, the amendment would ensure 
Recovery Act funds are expended by September 2017 and lead to reduced 
construction costs through reduced design-build contract bids. The 
original grant agreement called for a 50/50 match between federal and 
state funds. FRA officials said the grant amendment maintains this 
match but on a different time frame. 

[13] Department of Transportation Office of Inspector General, FRA 
Needs to Expand Its Guidance on High-Speed Rail Project Viability 
Assessments, CR-2012-083 (Washington, D.C.: Mar. 28, 2012). 

[14] GAO, High Speed Passenger Rail: Learning from Service Start-Ups, 
Prospects for Increased Industry Investment, and Federal Oversight 
Plans, [hyperlink, http://www.gao.gov/products/GAO-10-625] 
(Washington, D.C.: June 17, 2010). 

[15] Cal. Pub. Util. Code § 185033(a). 

[16] Cal. Pub. Util. Code § 185035(a), (c). 

[17] Cal. Sts. & High. Code § 2704.04(e). In April 2010 and January 
2012, the California state auditor released two reports identifying 
several risks with the project including funding uncertainties and 
concerns about the Authority's level of oversight of the project. 

[18] These revenue forecasts are in 2010 dollars. 

[19] These revenue forecasts are in 2011 dollars. 

[20] GAO, High-Speed Passenger Rail: Preliminary Assessment of 
California's Cost Estimates and Other Challenges, [hyperlink, 
http://www.gao.gov/products/GAO-13-163T] (Washington, D.C.: Dec. 6, 
2012). 

[21] [hyperlink, http://www.gao.gov/products/GAO-13-163T]. 

[22] The Authority developed a low cost estimate and a high cost 
estimate based on various potential routes for the train within the 
seven construction sections between San Francisco and Los Angeles. The 
Authority developed the low cost estimate based on a sum of the lowest 
cost route alignments in each construction section and, similarly, the 
high cost estimate reflects a sum of the highest cost route alignments 
in each construction section. The Authority then adjusted the 2011 
cost estimates to calculate year-of-expenditure, or nominal dollars, 
using assumptions on expected inflation rates. 

[23] [hyperlink, http://www.gao.gov/products/GAO-09-317]. 

[24] Examples of where we have applied the Cost Guide can be found in 
the following: GAO, Air Traffic Control Modernization: Management 
Challenges Associated with Program Costs and Schedules Could Hinder 
NextGen Implementation, [hyperlink, 
http://www.gao.gov/products/GAO-12-223] (Washington, D.C.: Feb. 16, 
2012); GAO, St. Lawrence Seaway: Estimates for the Asset Renewal 
Program Will Change, and Implementing Best Practices May Improve the 
Estimates' Reliability, [hyperlink, 
http://www.gao.gov/products/GAO-10-541R] (Washington, D.C.: May 13, 
2010); and, GAO, VA Construction: VA Is Working to Improve Initial 
Project Cost Estimates, but Should Analyze Cost and Schedule Risks, 
[hyperlink, http://www.gao.gov/products/GAO-10-189], (Washington, 
D.C.: Dec. 14, 2009). 

[25] The DOT OIG has since developed best practices for developing 
operating cost estimates for high-speed rail projects and has 
recommended that FRA update its guidance to HSIPR grantees. See DOT 
OIG, FRA Needs to Expand its Guidance on High-speed Rail Project 
Viability Assessments, CR-2012-083 (Washington, D.C.: Mar. 28, 2012). 

[26] The Authority is mandated by state law to include in its 2014 
business plan the results of a study by the UIC, examining how the 
estimated operating costs compares to rail systems in other countries. 

[27] The bids for the first construction project were not yet made 
public at the time this report was prepared. 

[28] The Authority was a project participant with MTC. The primary 
objective of this work was to provide information for the development 
of the Bay Area Regional Rail Plan, and to provide information to 
update environmental analyses to be conducted by the Authority. 

[29] [hyperlink, http://www.gao.gov/products/GAO-13-163T]. 

[30] [hyperlink, http://www.gao.gov/products/GAO-09-317]. 

[31] We further recommended that the methods could include such things 
as independent, third-party reviews of applicable ridership and other 
forecasts, identifying and implementing ways to structure incentives 
to improve the precision of ridership and cost estimates received from 
grant applications, or other methods that can ensure a high degree of 
reliability of such forecasts. 

[32] Department of Transportation OIG, FRA Needs to Expand Its 
Guidance on High-Speed Rail Project Viability Assessments, CR-2012-083 
(Washington, D.C.: Mar. 28, 2012). 

[33] Steer Davies Gleave, Ltd., prepared for the U.S. Department of 
Transportation, Office of Inspector General, High Speed Intercity 
Passenger Rail Best Practices: Ridership and Revenue Forecasting 
(April 2011). 

[34] Developing a ridership model involves many design decisions, 
assumptions, and inputs with varying degrees of impact on the 
resulting forecasts. Of necessity, we did not evaluate every design 
decision, assumption, and input, so our analysis was limited to a high-
level evaluation of the major steps taken. For the purposes of our 
analysis, we assessed "reasonableness" of the Authority's ridership 
and revenue forecasts by the extent to which the methodology used to 
generate these forecasts follows generally accepted practices for high-
speed rail ridership modeling, and as such should be expected to 
produce generally plausible and defensible model results. A reasonable 
assessment indicates a level of confidence in the methods used to 
generate the ridership forecast and the resulting forecasts for 
decision-making purposes. 

[35] Steer Davies Gleave, High Speed Intercity Passenger Rail Best 
Practices, April 2011. 

[36] Trip tables summarize data on the number of trips by mode for 
each city pair along the proposed route. 

[37] The 2005 survey was performed under contract to and under the 
direction of the MTC. The Authority was a "project participant" with 
the MTC. 

[38] The Authority subcontracted with Harris Interactive--a market 
research firm--to conduct the May 2011 long-distance travel survey. 
The survey data collected were used to verify whether travel 
characteristics had changed since the earlier survey. 

[39] The Authority used existing data from the Metropolitan 
Transportation Commission and Southern California Association of 
Governments. 

[40] R. Baker, et. al., Public Opinion Quarterly, Prepared for the 
American Association for Public Opinion Research, AAPOR Report on 
Online Panels, October 20, 2010, 74(4): 711-781. AAPOR is an 
association comprised of professionals working in the survey and 
public opinion research field on various topics such as election 
polling, market research, statistics, and research methodology. AAPOR 
produces work focused on developing, among other resources, best 
practices in collecting, analyzing, and interpreting survey data. 

[41] Caltrain is a commuter rail service on the San Francisco 
Peninsula. According to Authority officials, the purpose of this 
sensitivity analysis was to understand the impact of running less 
frequent services in the Peninsula in order to better address the 
capacity issues on the Caltrain corridor and potentially save capital 
costs. Two primary changes were made to service characteristics in 
this analysis. The first was a reduction in Caltrain service from 12 
trains per hour in the peak periods to 6 trains per hour, and only 1 
train per hour in the off-peak period. The second change was a 
modification to the high-speed rail operating plan whereby service 
from San Francisco Transbay Terminal to Los Angeles Union Station was 
reduced from 4 trains per hour to 2 trains per hour in the peak and 
off-peak periods. Service from San Jose to Los Angeles Union Station 
was increased from 2 trains per hour to 3 trains per hour in the peak 
period. According to Authority officials, a high-speed rail service 
frequency of 4 trains per hour in the Peninsula is the service 
alternative currently being considered by the Authority and is 
identified in the April 2012 revised business plan. A 2 trains per 
hour alternative has not been approved. 

[42] Academic experts from the University of California Berkeley's ITS 
conducted a review of ridership and revenue forecast models used to 
develop forecasts in June 2010 and produced a report summarizing their 
findings. See D. Brownstone, M. Hansen, and S. Madanat, Review of Bay 
Area/California High-Speed Rail Ridership and Revenue Forecasting 
Study, University of California Berkeley's Institute of Transportation 
Studies, UCB-ITS-RR-2010-1, June 2010. 

[43] Some decisions, such as parameter adjustments, are made based on 
professional judgment and experience of forecast modelers and were not 
evaluated as part of our review. 

[44] D. Brownstone, M. Hansen, and S. Madanat, Review of Bay Area/ 
California High-Speed Rail Ridership and Revenue Forecasting Study, 
University of California Berkeley's Institute of Transportation 
Studies, UCB-ITS-RR-2010-1, June 2010. 

[45] M. Bierlaire, D. Bolduc, and D. McFadden, The Estimation of 
Generalized Extreme Value Models from Choice-based Samples. 
Transportation Research Board (2008) 42. 

[46] In 2012, ridership on Amtrak's premium Acela service was nearly 
3.4 million passengers and total ridership on the regional NEC service 
(Acela and non-Acela) was more than 8.0 million passengers. 

[47] For example, the populations of cities along the NEC range from 
0.3 to 18.8 million while the populations of cities along the Los 
Angeles to San Francisco corridor range from 0.1 to 12.9 million. 

[48] [hyperlink, http://www.gao.gov/products/GAO-09-317]. 

[49] Bent Flyvbjerg, "Quality Control and Due Diligence in Project 
Management: Getting Decisions Right by Taking the Outside View," 
International Journal of Project Management (Nov. 2012). 

[50] Federal funds for construction or improvement of federal-aid 
highways generally must be matched with funds from other sources. For 
example, unless otherwise specified in authorizing legislation, most 
projects that are part of the Interstate System have an 80 percent 
federal share. 23 U.S.C. § 120(b). 

[51] [hyperlink, http://www.gao.gov/products/GAO-09-317]. 

[52] The President's fiscal year 2013 budget request proposed 
providing multi-year funding ($47 billion over 6 years) for high-speed 
rail. 

[53] GAO, High-Risk Series: An Update, [hyperlink, 
http://www.gao.gov/products/GAO-13-283], (Washington, D.C.: February 
2013) 

[54] According to their April 2012 revised business plan, the 
Authority is planning on applying $8.2 billion of the $9.95 billion in 
state high-speed rail bond proceeds to project construction. Some $950 
million of the remaining $1.75 billion will be used for transportation 
projects that will connect to high-speed rail. And, the other $800 
million will be used for environmental, planning, and support costs. 

[55] According to the California Air Resources Board--the state agency 
in charge of implementing the cap-and-trade program--California law 
requires that the volume of greenhouse gas emissions generated in the 
state be lowered to 1990 levels by 2020. To meet this goal, California 
established a limit on annual greenhouse gas emissions, or, a cap. The 
cap on pollution is divided into allowances that have to be acquired 
by polluting entities, such as coal-burning power plants. The state 
conducts periodic auctions to sell these allowances and also allows 
them to be traded (bought and sold) on a secondary market. 

[56] According to a report from California's Legislative Analyst's 
Office, the November 2012 auction was expected to result in a price of 
$10 to $50 per credit. According to the California Air Resources 
Board, the November 2012 auction resulted in a price of about $10 and 
raised approximately $290 million in auction proceeds. 

[57] As we reported in February 2008, an operating concession is a 
type of public-private partnership that offers rights to a private 
party to operate in return for a fixed-fee or percentage of revenues. 
See GAO, Highway Public-Private Partnerships: More Rigorous Up-front 
Analysis Could Better Secure Potential Benefits and Protect the Public 
Interest, [hyperlink, http://www.gao.gov/products/GAO-08-44] 
(Washington, D.C.: Feb. 8, 2008). For the California high-speed rail 
project, an operating concession would be an agreement between the 
Authority and a private firm, or consortium of firms, to retain ticket 
revenues with the responsibility to operate and maintain the service, 
in exchange for an up-front payment to the Authority. 

[58] [hyperlink, http://www.gao.gov/products/GAO-09-317]. 

[59] The Authority is planning for approximately $38.7 billion in 
additional federal funds for completing the entire Phase 1 of the 
project. Of this, approximately $25 billion will be used to complete 
the Initial Operating Segment. Operating profits, which are total 
operating revenues less operating expenses, are necessary to make an 
operating concession viable without other guarantees of income to the 
private investor. 

[60] Intercity passenger rail service in the United States is provided 
by Amtrak mostly over tracks owned by private freight railroads. There 
have been recent proposals to use private operators other than Amtrak 
for new intercity passenger rail services, such as the Desert Xpress 
and All Aboard Florida. 

[61] [hyperlink, http://www.gao.gov/products/GAO-08-44]. 

[62] British pounds converted to U.S. dollars based on the November 
15, 2010 exchange rate according to historical data published by the 
Board of Governors of the Federal Reserve System. 

[63] The Authority's 2012 revised business plan estimates an average 
of $937 million in annual operating profits, from 2022 through 2034. 

[64] 74 Fed. Reg. 29900 (June. 23, 2009). 

[65] User impacts (or benefits) are the impacts on passengers who 
actually use a transportation improvement. That is, travelers must use 
the improvement to benefit from it. Non-user impacts (or benefits) are 
those impacts on those who do not directly use an improvement. 

[66] A benefit-cost analysis is designed to identify the alternative 
with the greatest net benefit by comparing the monetary value of 
benefits and costs of each alternative with a baseline. Benefit-cost 
analysis provides for a comparison of alternatives based on economic 
efficiency, that is, which investment or policy would provide the 
greatest net benefit (i.e., greater than costs). Economic impact 
analysis is a tool for assessing how the benefits and costs of 
transportation alternatives would be distributed throughout the 
economy and for identifying groups in society (for example, by region, 
income, or race) that are likely to gain from, or bear the costs of, a 
policy. 

[67] 42 U.S.C § 4321 et seq. 

[68] DOT OIG, FRA Needs to Expand Its Guidance on High-Speed Rail 
Project Viability Assessments, CR-2012-083 (March 28, 2012). In 
particular, we focused on best practices contained in the report HSIPR 
Best Practices: Public Benefits Assessment, Steer Davies Gleave, June 
2011 that was prepared for the Office of Inspector General in 
conjunction with the March 2012 report. The Steer Davies Gleave report 
included a list of components that would be included in a public 
benefits assessment of high-speed rail projects. The components of 
public benefit assessments were similar to economic impacts identified 
through a review of the HSIPR NOFAs. 

[69] GAO, Bus Rapid Transit: Projects Improve Transit Service and Can 
Contribute to Economic Development, [hyperlink, 
http://www.gao.gov/products/GAO-12-811] (Washington, D.C.: July 25, 
2012). Bus rapid transit projects vary in design but generally include 
service enhancements designed to attract riders and provide similar 
transit-related benefits to rail transit. For the purposes of the 
report the term economic development was used to refer to components 
of transit-oriented development, such as high-density, mixed-use 
developments and pedestrian-friendly environments and streetscapes. 

[70] GAO, Affordable Housing in Transit-Oriented Development: Key 
Practices Could Enhance Recent Collaboration Efforts between DOT-FTA 
and HUD, [hyperlink, http://www.gao.gov/products/GAO-09-871] 
(Washington, D.C.: Sept. 9, 2009). 

[71] GAO, Commuter Rail: Potential Impacts and Cost Estimates for the 
Cancelled Hudson River Tunnel Project, [hyperlink, 
http://www.gao.gov/products/GAO-12-344] (Washington, D.C.: Mar. 9, 
2012). 

[72] According to the Bureau of Labor Statistics, a full employment 
economy in 2020 would have a 5.2 percent unemployment rate. See Bureau 
of Labor Statistics, Monthly Labor Review, January 2012. 

[73] The BCA states that different scenarios were considered in 
developing the analysis period and, depending on the scenario, the 
construction period varied. The analysis period used was 47 years 
beyond project completion for the scenario with the longest 
construction period. The assumed operating period is not intended to 
indicate the system will stop operating in 2060. 

[74] According to the EIA, the unemployment rate over the San 
Francisco Bay Area to Los Angeles corridor varied. The EIA states that 
through the recent recession, the San Francisco Bay Area unemployment 
rate was 1.0 to 1.5 percentage points less than the state as a whole. 
At the time of the EIA, the Central Valley unemployment rate about 4 
percent higher than the statewide average and the Southern California 
unemployment rate was within 1 percent of the statewide rate. 

[75] The Authority's EIA states that a "job-year" refers to the idea 
that one person working one job for 20 years represents 20 job-years. 
It is not the number of individuals working. Also, the employment 
factor used to develop the estimate is for construction and capital 
spending and not employment that might be associated with spending for 
operation and maintenance of the high-speed rail system. 

[76] HSIPR Best Practices: Public Benefits Assessment, Steer Davies 
Gleave, June 2011, prepared for Office of Inspector General, U.S. 
Department of Transportation. 

[77] U.S. Department of Transportation, Federal Highway 
Administration, Economic Analysis Primer (August 2003). 

[78] GAO, Intercity Passenger and Freight Rail: Better Data and 
Communication of Uncertainties Can Help Decision Makers Understand 
Benefits and Trade-offs of Programs and Policies, [hyperlink, 
http://www.gao.gov/products/GAO-11-290] (Washington, D.C.: Feb. 24, 
2011); Executive Order 12893, Principles for Federal Infrastructure 
Investments(Jan. 26, 1994); OMB Circular A-94 Revised, Guidelines and 
Discount Rates for Benefit-Cost Analysis of Federal Programs (Oct. 29, 
1992); and OMB Circular A-4, Regulatory Analysis (Sept. 17, 2003). 

[79] [hyperlink, http://www.gao.gov/products/GAO-11-290]. 

[80] California High-Speed Rail Authority, Comparison of Providing 
Equivalent Capacity to High-Speed Rail through Other Modes (April 
2012). 

[81] The scope of the analysis was the 520-mile Phase 1 system. 
Comparisons between high-speed rail and highways and aviation were 
based on system capacity, or "people carrying" capacity, not ridership 
estimates. According to the analysis, system capacity was used, in 
part, to recognize that capacity provides a steady state comparison 
between transport modes since it is tied to physical infrastructure, 
not the number of people using that infrastructure. The Authority 
assumed 12 trains per hour would be operated in each direction, 1,000 
seats per train, and an average load factor of 70 percent. Load factor 
is the number of average passengers on a train divided by the number 
of available seats. 

[82] California Transportation Commission, 2011 Statewide 
Transportation System Needs Assessment, Final Report (November 2011). 

[83] GAO, Cost Estimating and Assessment Guide: Best Practices for 
Developing and Managing Capital Program Costs, [hyperlink, 
http://www.gao.gov/products/GAO-09-3SP] (Washington, D.C.: March 2009). 

[84] For the purposes of our analysis, we assessed "reasonableness" of 
the Authority's ridership and revenue forecasts by the extent to which 
the methods used to generate these forecasts followed generally 
accepted practices for high-speed rail demand modeling, and as such 
should be expected to produce generally plausible and defensible model 
results. A reasonable assessment indicates a level of confidence in 
the methods used to generate the ridership forecast and the resulting 
forecasts for decision-making purposes. 

[85] Steer Davies Gleave, prepared for the U.S. Department of 
Transportation, Office of Inspector General, High-Speed Intercity 
Passenger Rail Best Practices: Ridership and Revenue Forecasting 
(April 2011). 

[End of section] 

GAO’s Mission: 

The Government Accountability Office, the audit, evaluation, and 
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 GAO’s website [hyperlink, http://www.gao.gov]. Each 
weekday afternoon, GAO posts on its website newly released reports, 
testimony, and correspondence. To have GAO e-mail you a list of newly 
posted products, go to [hyperlink, http://www.gao.gov] and select 
“E-mail Updates.” 

Order by Phone: 

The price of each GAO publication reflects GAO’s actual cost of 
production and distribution and depends on the number of pages in the 
publication and whether the publication is printed in color or black 
and white. Pricing and ordering information is posted on GAO’s 
website, [hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or 
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card, 
MasterCard, Visa, check, or money order. Call for additional 
information. 

Connect with GAO: 

Connect with GAO on facebook, flickr, twitter, and YouTube.
Subscribe to our RSS Feeds or E mail Updates. Listen to our Podcasts.
Visit GAO on the web at [hyperlink, http://www.gao.gov]. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 
Website: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]; 
E-mail: fraudnet@gao.gov; 
Automated answering system: (800) 424-5454 or (202) 512-7470. 

Congressional Relations: 

Katherine Siggerud, Managing Director, siggerudk@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, DC 20548. 

Public Affairs: 
Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, DC 20548. 

[End of document]