This is the accessible text file for GAO report number GAO-11-134 
entitled 'Surface Freight Transportation: A Comparison of the Costs of 
Road, Rail, and Waterways Freight Shipments That Are Not Passed on to 
Consumers' which was released on February 28, 2011. 

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 the Subcommittee on Select Revenue Measures, Committee on 
Ways and Means, House of Representatives: 

January 2011: 

Surface Freight Transportation: 

A Comparison of the Costs of Road, Rail, and Waterways Freight 
Shipments That Are Not Passed on to Consumers: 

GAO-11-134: 

GAO Highlights: 

Highlights of GAO-11-134, a report to the Subcommittee on Select 
Revenue Measures, Committee on Ways and Means, House of 
Representatives. 

Why GAO Did This Study: 

Road, rail, and waterway freight transportation is vital to the nation’
s economy. Government tax, regulatory, and infrastructure investment 
policies can affect the costs that shippers pass on to their 
customers. If government policy gives one mode a cost advantage over 
another, by, for example, not recouping all the costs of that mode's 
use of infrastructure, then shipping prices and customers' use of 
freight modes can be distorted, reducing the overall efficiency of the 
nation’s economy. 

As requested, this report (1) describes how government policies can 
affect competition and efficiency within the surface freight 
transportation sector, (2) determines what is known about the extent 
to which all costs are borne by surface freight customers, and (3) 
discusses the use of the findings when making future surface freight 
transportation policy. GAO reviewed the transportation literature and 
analyzed financial and technical data from the Department of 
Transportation (DOT), the Army Corps of Engineers (Corps), and the 
Environmental Protection Agency to make cross-modal comparisons at a 
national level. Data limitations and assumptions inherent in an 
aggregate national comparison are noted in the report. 

GAO is not making recommendations in this report. GAO provided a draft 
of this report to DOT and the Corps. DOT provided technical 
suggestions and corrections, which were incorporated as appropriate. 
The Corps had no comments. 

What GAO Found: 

Public spending, tax, and regulatory policies can promote economic 
efficiency in the freight transportation sector when they result in 
prices that reflect all marginal costs (the cost to society of one 
additional unit of service). These costs include private costs; public 
costs, such as infrastructure maintenance; and external costs, such as 
congestion, pollution, and accidents. When prices do not reflect all 
these costs, one mode may have a cost advantage over the others that 
distorts competition. As a consequence, the nation could devote more 
resources than needed to higher cost freight modes, an inefficient 
outcome that lowers economic well-being. Inefficient public investment 
decisions can result when all construction and other fixed costs are 
not passed on to the beneficiaries of that investment. 

GAO’s analysis shows that on average, additional freight service 
provided by trucks generated significantly more costs that are not 
passed on to consumers of that service than the same amount of freight 
service provided by either rail or water. GAO estimates that freight 
trucking costs that were not passed on to consumers were at least 6 
times greater than rail costs and at least 9 times greater than 
waterways costs per million ton miles of freight transport. Most of 
these costs were external costs imposed on society. Marginal public 
infrastructure costs were significant only for trucking. Given 
limitations in the highway, rail, and waterway economic, financial, 
technical, and environmental data available for the analysis, GAO 
presents conservative estimates. 

While freight costs are not fully passed on to consumers across all 
modes, a number of issues are important for decision makers to 
consider when proposing policy changes to align prices with marginal 
costs or reduce the difference between government fixed costs and 
revenues. Costs can vary widely based on the specific characteristics 
of an individual shipment, such as the geography and population 
density of the shipment’s route, and the fuel-efficiency of the 
specific vehicle carrying it. Policy changes that align prices with 
marginal costs on a shipment-by-shipment basis would provide the 
greatest economic benefit, but precisely targeted policy changes can 
result in high administrative costs. By contrast, less targeted 
changes—such as charging user fees based on average costs, subsidizing 
more efficient alternatives, or broadly applying safety or emissions 
regulations—can change the overall distribution of freight across 
modes, but may provide fewer benefits. Although the current 
configuration of transportation infrastructure can limit the shifting 
of freight among modes, price changes can prompt other economic 
responses. Over the longer term, there is greater potential for 
responses that will shape the overall distribution and use of freight 
services. 

View [hyperlink, http://www.gao.gov/products/GAO-11-134] or key 
components. For more information, contact Phillip R. Herr at (202) 512-
2834 or herrp@gao.gov, or James R. White at (202) 512-9110 or 
whitej@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

Public Policies That Encourage Pricing Freight Transport at Levels 
That Reflect Social Costs Would Maximize Economic Well-Being, but 
Other Objectives Also Matter: 

Available Data Indicate That Consumers Do Not Pay the Full Costs of 
Transporting Freight, Particularly Freight Moved by Truck: 

Formulating Policy Responses to Address Unpriced Social Costs Raises 
Complex Issues: 

Agency Comments: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: Recovery Act Funds for Freight Transportation 
Infrastructure: 

Appendix III: Federal Tax Subsidies and Financing Programs: 

Appendix IV: Freight External Cost Estimates from the Literature: 

Appendix V: Policy Options for Addressing Tradeoffs between Efficiency 
and Cost Recovery: 

Appendix VI: GAO Contacts and Staff Acknowledgments: 

Tables: 

Table 1: Estimated Ton-Miles of Domestic Surface Freight Shipped by 
Mode in 2007: 

Table 2: Categories of Total Social Costs in the Freight 
Transportation Sector: 

Table 3: Estimates of Marginal Social Costs Attributable to Each 
Freight Mode Not Passed on to Consumers, per Million Ton-Miles: 

Table 4: Cross-Modal Comparisons of Externalities: 

Table 5: Estimates of Fixed Social Costs Attributable to Each Freight 
Mode That Are Not Passed on to Consumers, per Million Ton-Miles: 

Table 6: Estimated Average Infrastructure Expenditures and Revenue 
(per Million Ton-Miles) by Level of Government by Mode: 

Table 7: Estimated Truck Ton-Miles of Domestic Surface Freight, 2000 - 
2007: 

Table 8: Methodology for Estimating Average Annual Accident Fatalities 
and Injuries, per Billion Ton-Miles, Average of 2003 to 2007: 

Table 9: Methodology for Estimating Tons of Freight-Related PM2.5 and 
NOX Emissions, per Million Freight Ton-Miles for Trucks and 
Locomotives in 2002 and for Waterborne Vessels in 2005: 

Table 10: Methodology for Estimating Damages of Freight-Related PM2.5 
and NOX Emissions, per 2002 Ton-Miles for Trucks and Locomotives and 
2005 Ton-Miles for Waterborne Vessels: 

Table 11: Methodology for Estimating Freight-Related CO2 Emissions, 
per 2007 Ton-Miles for Trucks and Locomotives and 2005 Ton-Miles for 
Waterborne Vessels: 

Table 12: Summary of Recovery Act Funds for Transportation 
Infrastructure Projects That Might Benefit Specific Modes: 

Table 13: Summary of Federal Tax Subsidies and Financing Programs, 
Fiscal Years 2003-2007: 

Figures: 

Figure 1: Tonnage of Freight on Highways, Railroads, and Inland 
Waterways (2007): 

Figure 2: Hypothetical Scenarios Illustrating How Prices That Do Not 
Reflect Social Costs Affect Resource Use and Competition: 

Abbreviations: 

BTS: Bureau of Transportation Statistics: 

CBO: Congressional Budget Office: 

CO2: carbon dioxide: 

Corps: U.S. Army Corps of Engineers: 

DOT: Department of Transportation: 

EPA: Environmental Protection Agency: 

FAF: Freight Analysis Framework: 

FHWA: Federal Highway Administration: 

NOX: nitrogen oxides: 

PM2.5: fine particulate matter with a diameter of 2.5 microns or less: 

Recovery Act: The American Recovery and Reinvestment Act of 2009: 

TIGER: Transportation Investment Generating Economic Recovery: 

TRB: Transportation Research Board: 

VMT: Vehicles Miles Traveled: 

[End of section] 

United States Government Accountability Office: 
Washington, DC 20548: 

January 26, 2011: 

The Honorable Patrick J. Tiberi: 
Chairman: 
The Honorable Richard E. Neal: 
Ranking Member: 
Subcommittee on Select Revenue Measures: 
Committee on Ways and Means: 
House of Representatives: 

Freight shipments move over vast networks of highways, railroads, and 
waterways--often transported by more than one mode before reaching 
their final destination.[Footnote 1] These networks connect and 
intersect, and play a critical role in providing the American public 
with the freight mobility needed to sustain national economic vitality 
and international competitiveness. According to the Department of 
Transportation (DOT), our surface freight transportation system 
connects an estimated 8 million businesses and 116 million households 
moving $12 trillion in goods.[Footnote 2] The movement of goods 
involves both private and public interests from private trucking 
companies, railroads, and waterborne vessel operators to federal, 
state, and local governments. While the major freight railroad 
infrastructure is privately owned and operated, and port 
infrastructure is privately or publicly owned and operated, 
governments play a primary role in planning, building, maintaining, 
and operating highways and keeping our waterways navigable. This 
infrastructure is designed for multiple types of users, not just 
freight service providers. Governments collect taxes and tolls, which 
help offset transportation expenditures, but have a minimal role in 
the direct regulation of prices and rates. Governments also regulate 
various aspects of freight transportation across all modes, including 
pollution, safety, and, to a more limited degree, congestion. 

How governments tax, regulate, and make investment decisions across 
modes could affect relative freight shipping prices. If government 
policy results in giving one mode of freight transportation a cost 
advantage over others--by, for example, ensuring that the wear and 
tear costs on infrastructure from users are fully recouped in one 
mode, but not in another mode--then shipping prices and choices made 
between alternative shipping options could be distorted. As a 
consequence, freight may be moved by a mode--for some portion or all 
of a trip--that imposes higher costs on the general public than might 
occur if such distortions did not exist. Because of your interest in 
the potential impact these policies can have on the freight 
transportation sector and beyond, you asked us to (1) describe how 
such policies can affect competition and efficiency within the surface 
freight transportation sector; (2) determine what is known about the 
extent to which costs are borne by surface freight users; and (3) 
discuss how our findings could be used when making future surface 
freight transportation policy. 

To address these objectives, we reviewed reports issued by the 
Congressional Budget Office (CBO), DOT, the Transportation Research 
Board (TRB), and the Brookings Institution. We interviewed officials 
from DOT, the Army Corps of Engineers (Corps), the Environmental 
Protection Agency (EPA), representatives from professional research 
organizations and industry, members of academia, and individuals 
knowledgeable about freight transportation to obtain advice on 
economic concepts, appropriate and available data sources, 
methodological approaches, and views on government spending and 
regulatory policies. We obtained preliminary reviews about the scope, 
methodology, and analysis contained in this report from DOT, EPA, the 
Corps, as well as two members of the Comptroller General's Advisory 
Board--comprised of individuals with broad expertise in public policy. 

We obtained, reviewed, and analyzed several datasets that can be used 
to estimate the revenues received from and costs imposed by users of 
the surface freight transportation system--federal, state, and local 
highways and roads; all classes of rail lines; and the inland, 
coastal, and Great Lakes waterways. Specifically, we identified data 
on federal, state, and local government revenues and expenditures on 
highways, railroads, and waterways from fiscal years 2000 through 
2006, the time frame of Federal Highway Administration's (FHWA) 
ongoing highway cost allocation study. We also obtained available data 
on external costs associated with freight transport, including 
pollution, accidents, and congestion from EPA, related research from 
DOT, and the Texas Transportation Institute. We analyzed these data to 
estimate the costs at a national level that freight users impose on 
the public transportation infrastructure and society and the revenues 
collected to offset those costs. When multiple data sources were 
available for our analyses, we explain why we selected one over 
another. 

To assess the reliability of the financial and technical data 
collected and published by various federal government agencies--such 
as DOT's Highway Statistics Series; the Department of the Treasury's 
statistics on income, debt, and tax expenditures; the Corps' 
Waterborne Commerce Statistics; and Census Bureau statistics, among 
others--we reviewed relevant documentation about the agencies' data 
collection and quality assurance processes, talked with knowledgeable 
officials from the agencies about these data, and compared these data 
against other sources of published information to determine data 
consistency and reasonableness. We determined that the data were 
sufficiently reliable for the purposes of providing high-level cost 
and revenue estimates by mode. 

We used federal statistical databases to obtain federal, state, and 
local data for estimating government costs and revenues. We also used 
nationwide data to estimate external costs. We recognize several 
important limitations in our high-level analysis, such as 
discrepancies in ton-mile estimates and difficulties in allocating 
costs between freight and nonfreight users. However, we explain how we 
deal with such limitations throughout the report by, for example, 
conducting sensitivity analyses to understand changes in costs with 
respect to ton-miles. Further, this analysis of high-level data is 
limited in the sense that it can obscure variations in state spending 
and revenue policies, and external costs by geographical location or 
by type of truck, locomotive, or marine engine. Moreover, this type of 
high-level analysis does not consider how modes compete with one 
another or the services or operations within each mode that compete 
with one another (e.g., rail long-haul with long-distance trucking); 
nor does it consider the complementary nature of freight modes, 
wherein, freight moved by rail or waterways may also involve trucks 
for at least some portion of its overall journey. Consequently, the 
results should be viewed as representing averages across all of the 
marginal shipments that were made under a wide variety of different 
conditions in a wide variety of locations. The last section of this 
report discusses the limitations that such high-level analyses have on 
policy evaluations. Appendix I details our objectives, scope, and 
methodology. 

We conducted our review from August 2009 to January 2011 in accordance 
with generally accepted government auditing standards. Those 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: 

Overview of the Surface Freight Transportation Sector: 

The nation's transportation infrastructure consists of over 4 million 
miles of public highways and roads; over 140,000 miles of national, 
regional, and local railroad networks; and 25,000 miles of 
commercially navigable waterways over which trillions of dollars worth 
of freight move annually. Public roads account for the majority of our 
nation's transportation infrastructure mileage, reaching nearly every 
corner of the United States, and as a result, enable trucks to move 
the greatest amount of freight on a tonnage basis. However, tonnage as 
a measure does not capture important aspects of freight mobility 
across the modes, such as the distances over which freight moves. For 
making comparisons across the modes throughout this report, we use ton-
miles as a unit of measurement. Ton-miles measure the amount of 
freight moved, as well as the distance over which it moves.[Footnote 
3] Table 1 shows the estimates and sources for ton-miles of freight 
moved on each mode for 2007, the most recent year that data are 
available. Appendix I provides more detail on our methodology for 
determining ton-miles used for the estimates in this report. 

Table 1: Estimated Ton-Miles of Domestic Surface Freight Shipped by 
Mode in 2007: 

Mode: Trucking; 
Ton-miles (in millions): 2,040,000; 
Source: Federal Highway Administration (FHWA), Freight Analysis 
Framework. 

Mode: Railroad; 
Ton-miles (in millions): 1,819,633; 
Source: Bureau of Transportation Statistics, National Transportation 
Statistics. 

Mode: Waterways; 
Ton-miles (in millions): 553,151; 
Source: U.S. Army Corps of Engineers, Waterborne Commerce of the 
United States. 

Sources: DOT and the Corps as indicated. 

[End of table] 

Freight shipments can also move by more than one mode before reaching 
their final destination. In particular, freight moved by rail or 
waterways may also be moved by truck at some point to reach its final 
destination, as rail and waterways may not reach locations that can be 
reached by truck. On the other hand, trains and waterborne vessels 
typically have far greater capacity than does a single freight truck, 
so rail and waterways generally move large volumes of commodities 
(e.g., coal and grain) long distances that would not be feasible by 
truck alone. Modes often work as complements to complete a shipment. 
For example, a ton of grain may move from a grain elevator by rail, be 
transported to a port on an inland waterway, move by barge to another 
port on an inland waterway, and then be distributed by truck to its 
final destinations. A particular type of shipment known as 
"intermodal" is designed to move on multiple modes, using a container 
that can be moved from a truck to a train to a ship without handling 
any of the freight itself when changing modes. Such freight movements 
are growing and FHWA forecasts that intermodal freight will continue 
to increase in the future.[Footnote 4] 

In some cases, the modes may be substitutable for certain types of 
trips and will compete directly for shipments or for segments of 
shipments based on price and performance. For example, long-haul 
trucking and rail shipments may be substitutable, or short sea 
shipping legs can be a substitute for rail or truck shipments along 
coastal routes. The extent to which mode-shifting is possible in the 
United States is difficult to estimate and will largely be determined 
by the types of parameters discussed above, such as whether shipping 
is feasible by another mode (e.g., rail lines or waterways may not be 
available for some routes), or practical (e.g., sending heavy coal 
shipments by truck or time-sensitive shipments by rail or waterways 
are not practical), and by the relative prices and other service 
characteristics of shipping by different modes.[Footnote 5] Figure 1 
geographically depicts the national freight transportation 
infrastructure and tonnage of freight activity by mode, which provides 
a sense of the physical reach of each modal network. 

Figure 1: Tonnage of Freight on Highways, Railroads, and Inland 
Waterways (2007): 

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

The following are depicted on the map: 

Modes of freight transportation: 
Inland waterways; 
U.S. Class I railroad[A]; 
National Highway system. 

Annual freight tons per route: 
62,500,000; 
215,000,000; 
250,000,000. 

Source: Highways: U.S. Department of Transportation, Federal Highways 
Administration, Freight Analysis Framework, Version 3.1, 2010. Rail: 
Based on Surface Transportation Board, Annual Carload Waybill Sample 
and rail freight flow assignments done by Oak Ridge National 
Laboratory. Inland Waterways: US. Army Corps of Engineers (Corps), 
Annual Vessel Operating Activity and Lock Performance Monitoring 
System data, as processed for Corps by the Tennessee Valley Authority; 
and Corps, Institute of Water Resources, Waterbourne Foreign Trade 
Data, Water flow assignments done by Oak Ridge National Laboratory. 
U.S. Department of Transportation (map). 

[A] In 2008, Class I railroads in the United States are defined as 
having annual carrier operating revenues of $401 million or more; 
Class II railroads are those with revenues greater than $32 million 
but less than $401 for at least 3 consecutive years; and Class III 
railroads are those with less than $32 million. 

[End of figure] 

Government Plays a Key Role in the Surface Freight Transportation 
Infrastructure: 

Federal, state, and local governments each play a crucial role in 
planning, designing, constructing, and maintaining the highways and 
waterways infrastructure, as well as raising revenues for the highway 
and waterway portions of the surface transportation system. 
Governments also play a role in regulating the freight industry, which 
we address in the next section. FHWA, state departments of 
transportation, and local transportation organizations plan and fund 
new highway infrastructure and maintain existing highways. The Corps 
has the responsibility for construction, operation, and maintenance of 
the nation's waterway system. There is limited public sector funding 
for rail infrastructure. All Class I railroads, which comprise about 
91 percent of all railroad revenues, are privately owned and, as one 
of the most capital-intensive industries in the United States, make 
considerable investments in their own transportation networks. 

* Highway infrastructure. The federal government authorized over $190 
billion for the federal-aid highway program for fiscal years 2005 
through 2009.[Footnote 6] A small portion of this funding was 
specifically identified for surface freight transportation projects, 
including $25 million for the freight Truck Parking Facilities program 
and $30 million for the Freight Intermodal Distribution Pilot Grant 
program.[Footnote 7] For the most part, however, funding is provided 
for construction, reconstruction, restoration, and rehabilitation of 
roads that serve both freight and nonfreight users. Because the 
federal government's expenditures for highways are based, in part, on 
the user pay principle, the government collects taxes and fees, which 
flow into the Highway Trust Fund--historically, the principal 
mechanism for funding federal highway programs. The fund's highway 
account reported revenues of about $34 billion in fiscal year 2007--
mainly from fuel (diesel and gasoline) tax that constitutes the 
majority of revenues from both freight and nonfreight users, as well 
as a variety of taxes imposed on trucks used in freight movement, 
including a truck and trailer sales tax, a heavy vehicle use tax, and 
a tire tax.[Footnote 8] In the following year, 2008, the Highway Trust 
Fund held insufficient amounts to sustain the authorized level of 
funding, and partly as a result, we placed it on our list of high-risk 
programs.[Footnote 9] To cover the shortfall, from fiscal years 2008 
through 2010 Congress transferred a total of $34.5 billion in 
additional general revenues into the Highway Trust Fund, including 
$29.7 billion into the highway account.[Footnote 10] Consequently, 
highway funding shifted away from the contributions of highway users, 
breaking the link between highway taxes paid and benefits received by 
users. The American Recovery and Reinvestment Act of 2009 (Recovery 
Act) further augmented transportation spending using general fund 
revenues of about $48 billion, of which about 57 percent was 
identified for federal highway projects.[Footnote 11] 

State and local governments also invest in public highways and roads. 
Within the federal-aid highway program, the federal government is 
responsible for funding 80 to 100 percent of highway project costs, 
and state and local governments are responsible for the remainder of 
the costs. State governments spent about $36 billion on capital 
outlays and about $21 billion more on maintenance of state-
administered highways in 2007, while local governments spent 
approximately $69 billion on public roads. According to FHWA, state 
governments collected about $61 billion in user revenue, and local 
governments collected about $4 billion from a combination of fuel 
taxes, vehicle taxes and fees, and tolls. State and local governments 
supplement user revenue with general fund appropriations to support 
highway and road activities. 

* Railroad infrastructure. The federal government has helped improve 
public safety on freight railroad infrastructure by providing limited 
funds to states for railroad-highway grade crossings and grants for 
relocating railroad tracks away from urban centers. Since January 1, 
2007, freight railroads no longer pay federal fuel taxes, and there is 
no federal user fee specific to freight railroads. However, the 
federal government pays freight rail companies for intercity passenger 
train usage of the companies' railroad tracks to the extent that these 
costs are not recovered through passenger fares. Recently, the 
Recovery Act funded two discretionary grant programs, the 
Transportation Investment Generating Economic Recovery (TIGER) grant 
program at $1.5 billion and the High-Speed Intercity Passenger Rail 
program at $8 billion, both of which can provide capital investment in 
railroad infrastructure.[Footnote 12] Additional funding for these 
programs were made available through the 2010 appropriations for DOT, 
[Footnote 13] nearly $600 million for TIGER grants and $2.5 billion 
for the high speed rail program. Because these programs are new, they 
are not included in the scope of our analysis of government spending 
on freight transportation. 

Little systematic information is available about state programs and 
financial assistance for the freight railroad industry. A 1997 survey 
of state departments of transportation found 10 states with dedicated 
freight railroad budgets exceeding $1 million annually.[Footnote 14] A 
few states (e.g., Alabama, North Dakota, and Tennessee) tax fuel for 
locomotives, but this revenue is not always used for rail projects. 
Railroads also pay state and local property taxes on their 
infrastructure; the nation's major railroads paid $625 million in 
property taxes in 2008, according to the Association of American 
Railroads.[Footnote 15] 

* Waterway infrastructure. The Corps, under its civil works program, 
is responsible for planning, constructing, operating, and maintaining 
the nation's waterways used primarily by commercial vessels, as well 
as recreational and commercial passenger boats along some sections of 
the waterways. For fiscal year 2007, the Corps spent about $1.2 
billion to operate and maintain the inland waterways, as well as the 
nation's coastal harbors and channels (deep and shallow draft), and 
$686 million more for a variety of construction projects along inland 
waterways and coastal harbors and channels. For the same year, the 
Saint Lawrence Seaway Development Corporation budgeted about $33 
million for operations and maintenance activities and $1 million for 
construction activities. Much of these funds are from the Harbor 
Maintenance Trust Fund.[Footnote 16] 

The general fund pays for all of the Corps' operations and maintenance 
activities and one-half of the inland waterway construction costs for 
rehabilitating, modernizing, or replacing locks and dams.[Footnote 17] 
The other half comes from commercial waterway users that pay fuel 
taxes which flow into the Inland Waterways Trust Fund. The Inland 
Waterways Users Board and the National Academy of Public 
Administration have both reported on inefficiencies in the delivery of 
construction projects which have led to delays and cost escalation 
that have strained the trust fund and resulted in fewer and less-
beneficial projects being funded.[Footnote 18] Some waterborne vessels 
are exempt from the fuel tax, including oceangoing ships, passenger 
boats, recreational craft, or government vessels. Receipts totaled 
about $101 million, including excise taxes and interest on 
investments, in fiscal year 2007. In contrast, revenue for the Harbor 
Maintenance Trust Fund comes largely from an excise tax on imports 
imposed on commercial users of certain ports. The tax applies a second 
time to cargo that has already arrived at a U.S. port, but is 
transferred by barge or short-sea route to another U.S. port. 
Importers or shippers pay an amount equal to 0.125 percent of the 
value of the commercial cargo involved at the time of unloading. 
Exporters are exempted from the excise tax. In fiscal year 2007, this 
trust fund received about $1.4 billion from tax collections--including 
$68 million from domestic shippers, which is relevant to the scope of 
this study--and $154 million from interest on investments in U.S. 
treasury bonds. Harbor Maintenance Trust Fund revenues exceed 
expenditures, and in 2007 the Fund was carrying a balance of nearly $4 
billion, which has continued to grow. The federal government levies 
other fees, such as customs and agricultural quarantine inspection 
fees, on waterborne vessel operators and shippers to cover the costs 
of the inspection programs. 

State and local governments also provide funding to publicly owned 
ports and dock facilities on waterways for the purposes of 
construction, operation, and maintenance of commercial port 
facilities, including warehouses, cranes, and terminals; canals; 
harbors; and other public waterways, in addition to dredging of those 
waterways. State and local governments also impose a variety of fees, 
such as canal tolls, rents from leases, concession rents, and other 
charges for use of commercial or industrial water transport and port 
terminal facilities and related services. 

Government Plays a Key Role in Regulating the Surface Freight 
Transportation Sector: 

In addition to constructing, operating, and maintaining the 
infrastructure, governments regulate various aspects of the surface 
freight transportation sector. Federal regulations across all modes 
are focused on safety and the environment rather than economic 
regulation. For example, truck safety regulations include truck size 
and weight limits and restrictions governing interstate freight 
operations.[Footnote 19] For rail, Congress has recently directed the 
Secretary of Transportation to require that Class I railroads, and 
commuter or regularly scheduled intercity transportation providers, 
install positive train control systems to help reduce the risk of 
crashes.[Footnote 20] Freight railroads continue to be subject to 
pricing regulation in areas where shippers do not have an alternative 
mode for shipping goods. Waterways freight carriers and their 
employees must comply with federal regulations. Indeed, all three 
modes--trucks, railroads, and waterway vessels--are expected to comply 
with federal drug testing,[Footnote 21] security,[Footnote 22] and 
environmental regulations, including measures imposing new pollution 
standards to reduce sulfur in diesel fuel.[Footnote 23] Except as 
preempted by federal law, state and local governments may also 
establish regulations that affect freight transportation. 

Compliance with these regulations can impose costs on the freight 
industries. For example, new emissions regulations may result in 
costlier investments in new vehicles than would otherwise have 
occurred. At the same time, government regulations are often intended 
to help reduce the costs of freight movements on society by reducing 
emissions and improving safety. 

Public Policies That Encourage Pricing Freight Transport at Levels 
That Reflect Social Costs Would Maximize Economic Well-Being, but 
Other Objectives Also Matter: 

In a market economy, resources are allocated to their most efficient 
uses (meaning they produce the greatest net benefits to society) when 
the prices of goods and services reflect all of the costs entailed in 
producing those goods and services. More specifically, economic 
efficiency requires that the price of a good or service equals the 
marginal social cost (the cost to society of consuming one additional 
unit of the good or service). Governments can best promote economic 
efficiency in the freight transportation sector by minimizing 
subsidies that produce gaps between prices and marginal social costs 
and by correcting price gaps that can occur naturally in the market. 
However, policies that promote efficiency can conflict with other 
objectives of policymakers, such as covering the costs of government 
services and satisfying certain concepts of equity. 

Government Subsidies and External Costs Can Result in Differences 
between the Costs of Freight Transportation Services and the Prices 
Charged to Shippers: 

The total social costs of providing freight transportation services 
can be divided into three categories on the basis of who bears them. 
First, there are private costs, such as labor, equipment, and fuel 
that are typically paid directly by freight service providers. Freight 
rail infrastructure falls into this category, as it is mainly funded 
privately by the rail companies. Second, there are the costs of public 
investments and services, such as the construction, maintenance, and 
operations of highways and waterways.[Footnote 24] These public costs 
are paid out of government budgets and can be funded through a variety 
of general or targeted taxes and fees. Finally, there are "external" 
costs, such as congestion, accidents, and health and environmental 
damage caused by pollution that are generated while transporting 
freight, that are not paid for directly by either the service 
providers or by government. These external costs are imposed on other 
members of society who are directly affected by these externalities. 
[Footnote 25] Each of these cost categories can be divided further 
between marginal costs and fixed costs. As noted earlier, marginal 
costs are those associated with the production of additional units of 
service. In contrast, fixed costs, such as those associated with the 
initial construction of infrastructure, are incurred before any 
service can be provided; however, the production of additional units 
of service does not add to these costs. 

In order to remain in business, private companies need to set prices 
that not only will cover their private marginal costs, but that will 
also include a margin that provides a sufficient rate of return to be 
able to obtain needed investment funds from capital markets. In a 
competitive market economy, only private costs will be passed on in 
prices to the final consumers of freight services, unless government 
policies are designed to pass the public and external costs on to 
those consumers as well. Governments can recover the public costs that 
support freight transportation by imposing taxes or fees on freight 
service providers. Competitive market forces should lead service 
providers to pass the cost of these payments on to their customers in 
the same manner that private costs are passed on. If competitive 
pricing prevents a particular business from passing such costs on to 
its customers, it may not earn a sufficient rate of return to remain 
in business.[Footnote 26] To the extent that public costs are not 
covered by taxes or fees levied on freight providers or consumers, 
governments would be providing a subsidy to the industry, which is 
paid by other taxpayers. Governments can also attempt to make freight 
service consumers bear the external costs generated by service 
providers by imposing taxes or fees on those providers in proportion 
to the external costs that they generate. Again, these costs should be 
passed on to the customers or noncompetitive businesses will drop out 
of the market. Government regulation of pollution and other factors 
that generate external costs can be used in conjunction with taxes and 
fees to address those costs. 

When Prices Do Not Reflect All Marginal Costs, Competition Can Be 
Distorted and Economic Efficiency Reduced: 

The hypothetical scenarios in figure 2 illustrate how discrepancies 
between marginal social costs (plus a competitive return on 
investment) and prices, whether caused by government subsidies or by 
external costs, can distort competition and cause inefficient 
allocations of resources in the freight transportation sector. 
[Footnote 27] In the scenarios, a shipper has to choose between two 
transportation modes to ship a package. Except for price, the services 
provided by the two modes are equal in all respects, such as 
timeliness and reliability. In the first scenario, Mode B uses $125 in 
resources to ship the package; Mode A uses $100 in resources. Price 
accurately reflects costs incurred to provide the freight service for 
both modes. Looking to minimize expenses, the shipper makes the 
logical decision and chooses the less expensive option (Mode A). The 
freight service provider represented by Mode A is rewarded for 
providing service more efficiently than the competitor, and the $25 of 
resources that otherwise would have been used if the product were 
shipped by Mode B can be used more efficiently in other ways to 
produce benefits for society. 

Figure 2: Hypothetical Scenarios Illustrating How Prices That Do Not 
Reflect Social Costs Affect Resource Use and Competition: 

[Refer to PDF for image: illustration] 

Scenario 1: Prices reflect social costs resulting in efficient 
resource use and fair competition: 

Mode A: 
Costs of freight service provider $100; 
Price to shipper $100. 

Mode B: 
Costs of freight service provider $125; 
Price to shipper $125. 

Shipper bases choice on prices: Mode A; 
Mode A: Since the price equals the lowest total cost, the result is 
efficient for society. 

Scenario 2: A subsidy results in inefficient resource use and 
competitive distortions: 

Mode A: 
Costs of freight service provider $100; 
Price to shipper $100. 

Mode B: 
Costs of freight service provider $125; 
Subsidy: $50; 
Price to shipper $75. 

Shipper bases choice on prices: Mode B; 
Mode A: The result is inefficient because more resources than 
necessary are used to ship the good. 

Scenario 3: External costs result in inefficient resource use and 
competitive distortions: 

Mode A: 
Costs of freight service provider: $100; 
External costs: $0; 
Costs to society: $100; 
Price to shipper: $100 

Mode B: 
Costs of freight service provider: $75; 
External costs: $50; 
Costs to society: $125; 
Price to shipper: $75. 

Shipper bases choice on prices: Mode B; 
The result is inefficient because the total costs to society of 
shipping the goods are higher than necessary. 

Source: GAO. 

[End of figure] 

The second scenario in figure 2 shows the detrimental effects of a 
subsidy.[Footnote 28] In this scenario, the government provides a 
subsidy to Mode B, enabling it to charge a price that is $50 below its 
marginal costs. As in the first scenario, the shipper selects the 
lower-priced option; however, in this case the subsidy results in the 
service being provided by the higher-cost producer. As a result, $25 
of resources that otherwise could have been used to provide other 
societal benefits are not used efficiently. 

The third scenario in figure 2 shows how external costs can distort 
competition and reduce economic efficiency in a manner similar to 
government subsidies. In this scenario, Mode B generates $50 in 
external costs that are not reflected in the price charged to the 
shipper. The fact that these costs are not passed on to the shipper 
makes Mode B more competitive than it would be if it had to include 
those costs in the price. Consequently, the shipper chooses Mode B, 
despite the fact that society bears $25 more in costs than if the 
other mode had provided the service. 

When prices do not reflect marginal social costs, investment decisions 
are also distorted, potentially resulting in a misallocation of 
resources. Much like a freight service shipper whose primary concern 
is price, an investor that is primarily concerned with profit 
potential is not concerned with the social costs that a freight 
service provider generates if they do not affect the provider's net 
profit. Therefore, an investor looking to maximize his or her return 
will invest resources in the more profitable provider regardless of 
social costs. From an economywide perspective, this is a misallocation 
of resources because those investment resources could be used more 
efficiently if applied to another area in the economy that is more 
efficient. 

There Can Be a Tradeoff between Recovering Fixed Public Infrastructure 
Costs and Promoting the Efficient Use of Existing Infrastructure: 

Certain freight transportation costs, such as the construction of new 
infrastructure, are considered to be "fixed" (rather than marginal) in 
the sense that they do not increase as use of the infrastructure 
increases. As an example, the construction cost of a bridge is a fixed 
cost, but pavement wear is a marginal cost. In freight transportation, 
fixed costs to build infrastructure are generally large relative to 
the marginal costs of an additional vehicle trip on an uncongested 
highway. Consequently, if governments were to charge users only for 
the marginal costs of their use, they usually would not be able to 
recover the full costs of building much of the infrastructure. As 
private companies that own and invest in their own infrastructure, 
freight railroads must pass on fixed costs to customers in order to 
remain in business.[Footnote 29] However, once the infrastructure is 
in place, charging users a portion of the fixed costs each time they 
use the infrastructure (on top of a charge for any marginal costs they 
impose) can result in underutilization of the infrastructure. Appendix 
V outlines a number of different ways that governments can address 
this tradeoff between efficiency and cost recovery.[Footnote 30] The 
choice among these alternatives involves a political, rather than a 
strictly economic judgment. Table 2 categorizes how the various types 
of costs in the freight transportation sector can be passed on to 
freight service consumers. 

Table 2: Categories of Total Social Costs in the Freight 
Transportation Sector: 

Marginal social costs (increase with each freight shipment): 

Category: Private; 
Examples: 
* Fuel; 
* Labor of truck drivers and rail and vessel operators; 
* Vehicle and tire wear; 
Circumstances under which the costs would be passed on to freight 
service consumers: These costs that are paid directly by freight 
service providers will generally be passed on to consumers in 
competitive markets. Where monopoly conditions exist (as is the case 
in certain rail corridors), effective government price regulation can 
ensure that excessive rates are not charged. 

Category: Public; 
Examples: 
* Pavement wear; 
* Wear on waterway locks and dams as vessels pass; 
Circumstances under which the costs would be passed on to freight 
service consumers: Government taxes or fees that are based on factors 
associated with infrastructure wear (e.g., vehicle miles traveled, 
loaded vehicle weight, and axle configuration) and levied on service 
providers would be passed on to consumers in the same manner that 
private costs are. 

Category: External; 
Examples: 
* Health and environmental damage due to pollution; 
* Time costs due to congestion (in certain places, at certain times); 
Circumstances under which the costs would be passed on to freight 
service consumers: Government taxes or fees based on volumes of 
pollutants produced by specific freight vehicles under specific 
conditions and time-variant tolls charged for specific routes would be 
passed on to consumers in the same manner that private costs are. 
Government regulations can also be used in conjunction with taxes and 
fees to reduce the amounts of external costs generated to begin with. 

Fixed social costs (exist regardless of whether an additional shipment 
is made): 

Category: Private; 
Examples: 
* Initial construction of warehouses, depots, and rail lines; 
Circumstances under which the costs would be passed on to freight 
service consumers: The cost of financing this construction will be 
incorporated in the prices that providers charge to consumers under 
the same conditions as described for private marginal costs above. 

Category: Public; 
Examples: 
* Construction of new highway capacity and maintenance of current 
highway stock; 
* Construction of locks and dredging of waterway channels; 
Circumstances under which the costs would be passed on to freight 
service consumers: Several alternative types of taxes or fees can be 
used to pass these costs on to consumers, including those that 
allocate the costs across all freight providers based on the extent of 
their use of the infrastructure. 

Category: External; 
Examples: 
* Health and environmental damage due to pollution from construction 
equipment; 
Circumstances under which the costs would be passed on to freight 
service consumers: Charges based on the volumes of pollutant could be 
included in the construction costs that the government allocates 
across freight providers in the manner described above. 

Source: GAO. 

[End of table] 

Policies That Promote Economic Efficiency Can Conflict with Other 
Government Objectives: 

Government policies aimed at reducing gaps between prices and social 
costs in the freight transportation sector also support the benefit 
principle of equity--a widely accepted economic principle--but they 
can conflict with the "ability-to-pay" principle of equity (which 
holds that people should contribute to the cost of government in line 
with their financial resources) and other objectives important to 
policymakers. The benefit principle holds that government services 
should be financed by those who benefit from those services. In the 
case of transportation funding, motor fuel taxes adhere more closely 
to the benefit principle than does the income tax because fuel 
consumption is correlated with road use. However, motor fuel taxes are 
regressive, meaning that lower income individuals pay a greater share 
of their income toward these taxes than do higher income individuals. 
This regressivity can conflict with the ability-to-pay principle, 
unless compensating relief to lower income individuals is provided in 
other parts of the tax system. 

Other objectives may be important to policymakers, such as whether or 
not a policy can be administered cost effectively. For example, 
attempts to achieve a high level of precision in marginal cost pricing 
through taxes and fees carry with them an administrative burden, as we 
discuss later in this report. The administrative costs of implementing 
finely calibrated versions of a tax may outweigh any efficiency gains 
achieved through increased precision. Efficiency in the freight 
transportation sector depends on prices fully reflecting marginal 
costs on a shipment-by-shipment basis; however, subsidies and external 
costs can vary considerably from one shipment to another based on the 
geographic origin and destination, time of day, and other factors. 
Moreover, as we discuss below, considerable uncertainty exists in the 
valuation of many types of costs. 

Available Data Indicate That Consumers Do Not Pay the Full Costs of 
Transporting Freight, Particularly Freight Moved by Truck: 

The combination of tax, spending, and regulatory policies in the 
United States does not result in consumers of all three surface 
freight transportation modes bearing the full costs they impose on 
society, particularly truck freight. Available data indicate that each 
of the modes, in the aggregate, generates marginal costs in excess of 
their marginal revenue.[Footnote 31] Specifically, we estimate that 
freight trucking costs that were not passed on to customers were at 
least 6 times greater than rail costs and at least 9 times greater 
than waterways costs per million ton miles of freight transport. Most 
of these costs were external costs imposed on society. In particular, 
the modes generate external costs related to accidents and pollution 
that are not reflected in prices. Furthermore, available data also 
indicate that at the national level, the infrastructure costs (both 
marginal and fixed) attributable to commercial freight transported by 
trucks and over waterways exceed the revenue that these freight 
transportation providers pay governments to fund that infrastructure. 
The available data for the freight transportation networks and 
vehicles we examined show that both the marginal and fixed social 
costs that are not passed on to freight service consumers are greatest 
(per million ton miles of freight carried) for freight trucks and 
lowest for railroads. 

Consumers of Freight Services Pay Less of the Marginal Costs 
Associated with Trucking than with Railroads or Waterways: 

Although certain data limitations and difficulties in valuing 
important categories of costs prevent us from making definitive 
quantitative estimates of the nonprivate (i.e., public and external) 
marginal costs generated by an additional million ton-miles of freight 
service provided by each of the three transportation modes, we are 
able to present at least lower bound estimates of those costs and to 
compare the magnitudes of these costs across the three modes.[Footnote 
32] In a competitive economy, private costs such as payments for labor 
and fuel are generally passed on in prices to the final consumers of 
freight services; therefore, those costs did not need to be included 
in our estimation of costs that are not passed on. We are also able to 
estimate the amount of revenue that governments collect from highway 
taxes and fees, such as those on motor fuels and tires that are 
associated with marginal activity.[Footnote 33] (We use the payment of 
such taxes and fees as a measure of the extent to which governments 
have passed some of the nonprivate costs on to final consumers). The 
extent to which the nonprivate marginal costs exceed tax and fee 
payments indicates the extent to which some nonprivate marginal costs 
are not reflected in prices charged to freight consumers. We refer to 
this difference as "unpriced costs." The available evidence suggests 
that, on average, an additional million ton-miles of freight service 
provided by trucking[Footnote 34] generates significantly more 
unpriced costs than an additional million ton-miles of either freight 
rail or waterways service generates.[Footnote 35] We estimate that 
over $55,000 per million ton-miles of service in unpriced freight 
trucking costs were not passed on to consumers. In contrast, freight 
rail and waterways services imposed over $9,000 and over $7,000 in 
unpriced costs per million ton-miles, respectively. 

Table 3 summarizes the estimates of marginal social costs attributable 
to each freight mode not passed on to consumers, per million ton 
miles. The estimates we present for pollution and other external costs 
are based on conservative volume estimates and valuation approaches 
from the available literature. Moreover, we do not include cost 
estimates for carbon dioxide (CO2) emissions because of the 
considerable uncertainty surrounding such estimates. For these 
reasons, our bottom-line estimates for marginal social costs not 
passed on to consumers are likely to represent minimum values for 
those costs.[Footnote 36] 

Table 3: Estimates of Marginal Social Costs Attributable to Each 
Freight Mode Not Passed on to Consumers, per Million Ton-Miles: 

Monetary values (in thousands of constant 2010 dollars): 

Marginal social costs: 

Marginal private costs: 
Trucking: [A]; 
Railroad: [A]; 
Waterways: [A]. 

Marginal public infrastructure costs (e.g., pavement preservation 
costs)[D]: 
Trucking: $7,000; 
Railroad: -[B]; 
Waterways: -[C]. 

Other public subsidies--federal tax subsidies and financing 
programs[E]: 
Trucking: -; 
Railroad: -; 
Waterways: -. 

Marginal external costs[F]: Emissions of particulate matter and 
nitrogen oxide; 
Trucking: $44,000; 
Railroad: $8,000; 
Waterways: $6,000[G]. 

Marginal external costs[F]: Accidents; 
Trucking: $8,000; 
Railroad: $1,000; 
Waterways: -. 

Marginal external costs[F]: Congestion; 
Trucking: $7,000; 
Railroad: -; 
Waterways: Unknown. 

Marginal taxes and fees: 

Taxes and fees associated with marginal freight activity; 
Trucking: $11,000; 
Railroad: -; 
Waterways: [C]. 

Marginal social costs not passed on to consumers[F]: 

Unpriced costs--marginal social costs minus taxes and fees associated 
with marginal freight activity; 
Trucking: Over $55,000; 
Railroad: Over $9,000 (but less than trucking costs that are not 
passed on); 
Waterways: Over $6,000 (but less than trucking costs that are not 
passed on). 

Source: GAO analysis of DOT and EPA data, except as noted. 

[A] As explained above, private costs are generally passed on to 
consumers, so they do not need to be added into the estimation of 
costs that are not passed on. 

[B] "-" means less than .5. 

[C] Transportation Research Board, Paying Our Way, Estimating Marginal 
Social Costs of Freight Transportation, (1996), shows 0.03 cents per 
ton-mile plus 3 cents per ton per lock passage, converted to 2010 
dollars by GAO. Also, Congressional Budget Office, Paying for 
Highways, Airways, and Waterways: How Can Users Be Charged? (May 
1992), shows 0.06 per ton-mile, converted to 2010 dollars by GAO. 

[D] Infrastructure costs and taxes and fees represent averages of data 
from fiscal years 2000 through 2006. 

[E] We did not include any state or local government tax subsidies or 
financing program targeted at the freight modes. 

[F] These estimates likely understate total external costs because 
they do not cover all types of external costs--for example, we did not 
calculate costs for CO2 emissions--and the estimates for the included 
costs are likely to be conservative. The conclusion that unpriced 
costs for rail and waterways are lower than those for trucking is 
based on data relating to emission, accident, and congestion volumes. 
The data in table 4 indicate that the excluded costs are larger for 
trucking than for the other modes. 

[G] This estimate is for inland waterways freight only because 
comprehensive data were not available for other types of waterways. 

[End of table] 

Marginal Public Infrastructure Costs: 

Marginal public infrastructure costs--the second cost item in table 3--
relate to public highway spending attributable to miles driven by 
freight trucks (i.e., pavement preservation costs per million ton- 
miles). We estimate from recent FHWA data that trucks imposed an 
average marginal cost to pavement of $7,000 per million ton- 
miles.[Footnote 37] We also estimate from FHWA data that pavement 
preservation costs borne by all levels of government attributable to 
all single-unit and combination trucks (excluding pickup trucks) 
averaged about 6.1 cents per vehicle miles traveled (VMT).[Footnote 
38] The cost per ton-mile would increase with truck weight and 
decrease with the number of axles. The costs also varied by location 
(urban or rural), type of road surface, temperature, and other 
factors. When we compared single-unit and combination trucks using DOT 
data, we found that marginal revenues exceeded the marginal 
infrastructure costs by 4.8 cents per VMT for single-unit trucks and 
by 3.5 cents per VMT for combination trucks, meaning that both types 
of trucks pay more than their share of pavement preservation costs. 
[Footnote 39] Although marginal costs are difficult to estimate from 
available data, CBO along with TRB and the Brookings Institution have 
undertaken this effort and reported their results. Their reports, 
although dated by at least 15 years, remain the most pertinent and 
relevant to our study. 

Because railroads generally pay for their own infrastructure, 
governments spend little on railroad infrastructure. For waterways 
freight, marginal public infrastructure costs, as estimated by TRB and 
CBO, are relatively low because the costs of dredging channels are 
predominantly fixed, rather than marginal, and vary little with the 
amount of tonnage that passes through. Because the Recovery Act (2009) 
was enacted after the time frame of our analysis and was a one-time 
funding source, our analysis does not consider these funds. Appendix 
II contains more details on the Recovery Act funds identified for 
freight transportation infrastructure by mode. 

Other Public Subsidies: 

Federal tax and financing programs subsidize the surface freight 
transportation infrastructure used by trucks, railroads, and 
waterborne vessels. Although we could not determine what portion of 
these benefits is associated with marginal activity, trucking and 
waterways freight received indirect, public subsidies through 
infrastructure improvements financed by certain state and local 
government bonds, which earned interest that was not subject to 
federal income tax. Trucking, railroads, and waterways also benefited 
from federal loan and loan guarantee financing programs for 
infrastructure improvements at attractive terms.[Footnote 40] However, 
we determined that the subsidies from federal financing programs for 
each of the three modes were negligible on a per-million-ton-mile 
basis. See appendix III for additional information on federal income 
tax subsidies and the federal financing programs. 

While each of the modes may benefit from certain provisions of the 
federal corporate income tax, the effects of these benefits on the 
three modes are not included in table 3 because they relate to fixed 
costs, rather than marginal costs. For example, eligible Class II and 
III railroads may take federal business tax credits for rail track 
maintenance, eligible shipping companies may make tax deferred 
deposits into a capital construction fund, and all of the modes can 
benefit from accelerated depreciation for tax purposes (as do many 
other industries). CBO's estimates of federal corporate effective tax 
rates for 2002[Footnote 41]--the best available evidence of whether 
the overall corporate income tax system favors one mode relative to 
another, or relative to other industries--suggest that the federal 
corporate income tax may provide a slight advantage to waterways 
freight over the other two modes.[Footnote 42] CBO estimated that the 
effective tax rate on the category of assets that includes heavy 
trucks, truck trailers, and buses--the category closest to freight 
trucks investments--to be 18.2 percent.[Footnote 43] Further, CBO 
estimated the effective tax rate on investments in railroad 
infrastructure to be 20.1 percent and the rate on investments in 
railroad equipment to be 11.4 percent. When weighted by the amounts of 
assets in railroad infrastructure and railroad equipment, these two 
rates combine for an average effective tax rate on railroad 
investments of 18.1 percent. The closest asset category for waterways 
freight includes all investments in ships and boats. CBO estimated the 
effective tax rate on these investments to be 16.5 percent. These 
relative effects are on top of any benefits due to public 
infrastructure investments that trucking and waterways receive over 
railroads. The effective tax rates for all three modes are below the 
26.3 percent average effective tax rate for all corporations, 
indicating that all three modes are receiving better than average tax 
rates. 

Marginal External Costs: 

For all of the freight modes, external costs are large relative to 
public infrastructure costs. Our analysis of available data to 
quantify the levels of externalities in table 4 shows that freight 
trucking produces more air pollution, accidents, and congestion per 
million ton-miles than do the other modes. However, we recognized that 
there are many difficulties in estimating the monetary costs 
associated with these external effects. Consequently, the estimates we 
presented previously in table 3 should be considered a rough order of 
magnitude estimate for these external costs.[Footnote 44] 

Table 4: Cross-Modal Comparisons of Externalities[A]: 

Category: Air pollution[C]; 
Type: Tons of particulate matter per million ton-miles, 2002; 
Trucking: 0.1191; 
Railroad: 0.0179; 
Waterways: 0.0116[D]; 
Trucking to rail ratio[B]: 6.7; 
Trucking to waterways ratio: 10.2. 

Category: Air pollution[C]; 
Type: Tons of nitrogen oxide per million ton-miles, 2002; 
Trucking: 3.0193; 
Railroad: 0.6747; 
Waterways: 0.4691[D]; 
Trucking to rail ratio[B]: 4.5; 
Trucking to waterways ratio: 6.4. 

Category: Air pollution[C]; 
Type: Tons of CO2 equivalents per million ton-miles, 2007; 
Trucking: 229.8; 
Railroad: 28.96; 
Waterways: 17.48; 
Trucking to rail ratio[B]: 7.9; 
Trucking to waterways ratio: 13.1. 

Category: Accidents[E]; 
Type: Fatalities per billion ton-miles, avg. 2003-2007; 
Trucking: 2.54; 
Railroad: 0.39; 
Waterways: 0.01; 
Trucking to rail ratio[B]: 6.4; 
Trucking to waterways ratio: 208.8. 

Category: Accidents[E]; 
Type: Injuries per billion ton-miles, avg. 2003-2007; 
Trucking: 55.98; 
Railroad: 3.32; 
Waterways: 0.05; 
Trucking to rail ratio[B]: 16.9; 
Trucking to waterways ratio: 1,239.6. 

Category: Congestion[F]; 
Type: Cost of delay to road users in 2000, (in billions of constant 
2010 dollars); 
Trucking: $10.86 billion; 
Railroad: $0.58 billion; 
Waterways: Not available; 
Trucking to rail ratio[B]: 18.6; 
Trucking to waterways ratio: Not available. 

Source: GAO analysis of data from DOT, EPA, and the Texas 
Transportation Institute. 

[A] Federal Highway Administration, Freight Facts and Figures 2009; 
and Bureau of Transportation Statistics, National Transportation 
Statistics. 

[B] A ratio of 1.0 indicates equal amounts of negative effect per unit 
of freight moved. For example, the ratio of 6.7 in the table indicates 
that truck freight produces, on average, six and seven-tenths times 
the particulate matter emissions as movement of the same unit of 
freight by rail. 

[C] Environmental Protection Agency, National Emissions Inventory, 
data provided to GAO by correspondence, and Inventory of U.S. 
Greenhouse Gas Emissions and Sinks: 1990 - 2008. 

[D] Estimate is for inland waterways freight only because 
comprehensive data were not available. Emissions data for waterways 
freight are for 2005 and were obtained from the Texas Transportation 
Institute, A Modal Comparison of Domestic Freight Transportation 
Effects on the General Public. 

[E] Federal Motor Carrier Safety Administration, Large Truck and Bus 
Crash Facts 2007; Federal Railroad Administration, Office of Safety 
Analysis online accident/incident data; and Federal Highway 
Administration, Freight Facts and Figures 2009. Trucks are defined as 
over 10,000 gross vehicle weight, which can include some non-freight 
activity. For example, in 2007, 12.3 percent of large trucks involved 
in a fatal accident and 13.2 percent in accidents with injuries were 
dump, garbage, or concrete-mixer trucks. 

[F] Federal Highway Administration, 1997 Federal Highway Cost 
Allocation Study Final Report. 

[End of table] 

Emissions and Air Pollution: 

EPA and DOT have not produced recent estimates of the economic costs 
of air pollution on a ton-mile basis for any of the freight modes. 
[Footnote 45] Therefore, we applied EPA's estimates for the human 
health benefits of reducing one ton of fine particulate matter and one 
ton of nitrogen oxide to the emissions data. We estimated for freight 
trucking an emissions cost of $44,000 per million ton-miles, as shown 
in table 3.[Footnote 46] Given the even greater uncertainty 
surrounding the economic costs of CO2 emissions, we did not produce 
our own estimate. The omission of these costs, as well as the omission 
of other nonhealth costs associated with emissions of nitrogen oxide 
and particulate matter, means that the estimates in table 3 are likely 
to understate the extent to which some marginal costs are not passed 
on to final consumers. This understatement would be the greatest for 
trucking. 

According to our synthesis of EPA's latest national emissions 
inventory data (2002), freight trucks produced over six times more 
fine particulate matter and over four times more nitrogen oxide on a 
ton-mile basis than freight locomotives,[Footnote 47] and over 10 and 
six times more of each type of emission, respectively, on a ton-mile 
basis than inland waterway vessels.[Footnote 48] And, according to our 
analysis of EPA data on greenhouse gases, trucks emitted the highest 
levels of greenhouse gas (CO2 equivalents) among the freight modes-- 
about eight times more per unit of freight than freight rail, and 
thirteen times more than waterways freight, as shown in table 4. 
[Footnote 49],[Footnote 50] Recent EPA regulatory changes require that 
freight carriers for all the modes upgrade to technologies that reduce 
particulate matter and nitrogen oxide emissions.[Footnote 51] EPA 
expects these standards to reduce diesel engine emissions of 
particulate matter and nitrogen oxide by 80 and 90 percent, 
respectively, for locomotives and waterborne vessels and 90 and 95 
percent, respectively for heavy duty trucks over the next 20 to 30 
years as older engines are taken out of service. While these 
regulations are expected to reduce the overall level of air pollution 
external costs, overall emissions will not be reduced to the estimated 
levels until 2030 or later because older, more polluting diesel 
engines will still be in use for years to come as each mode's fleet 
converts to the new technology. 

Accidents: 

According to our analysis of DOT data shown in table 4, nationwide 
between 2003 and 2007, large trucks were involved in about six times 
more accidents with fatalities and 17 times more accidents with 
injuries, per billion ton-miles, than freight rail. Rates of 
fatalities and injuries involving a waterways vessel were much lower 
than those involving both trucks and freight rail.[Footnote 52] The 
economic costs of transportation accidents reflect the value assigned 
to the loss of a human life and the reduced productive life and pain 
and suffering related to serious injuries.[Footnote 53] The external 
portion of those costs excludes any amounts borne by the freight 
service providers (e.g., through insurance premiums or court 
settlements). Available cost estimates from the literature, shown in 
appendix IV, indicate that truck external accident costs could be as 
much as 2.15 cents per ton-mile, almost nine times higher than rail 
external accident costs. However, these estimates are dated and do not 
reflect the reduced rate of truck and rail accidents in recent years, 
or the much higher economic value now assigned to loss of human life. 
To obtain our conservative estimate of $8,000 per million ton-miles in 
table 3, we started with the number of fatalities in table 4, 
multiplied by the latest value for human life used by DOT in guidance 
for its own analysts, and then assumed that carriers are already 
compensated for 50 percent of these costs (see appendix I for details 
on our scope and methodology). We identified four studies that 
attempted to determine the extent to which accident costs were 
compensated through insurance premiums, payments, and other 
compensation.[Footnote 54] These studies ranged from 50 to 62 percent 
in uncompensated or external costs. We chose to use 50 percent of the 
portion of costs that were not compensated as a reasonably 
conservative estimate since our calculations do not include estimates 
for uncompensated costs for injuries and property damage. 

Congestion: 

Most of the available information on road congestion, in particular 
the costs of delay for all highway users, does not specify external 
costs associated with freight traffic.[Footnote 55] We found only one 
study that provides a cross-modal estimate of freight congestion costs 
nationally, indicating that in 2000, congestion delay costs from 
intercity freight trucking were approximately five times those of 
intercity rail freight, per ton-mile.[Footnote 56] In its 1997 Highway 
Cost Allocation Study, FHWA estimated that in 2000 trucks were 
responsible for $10.9 billion (constant 2010 dollars) in congestion 
costs to other highway users nationwide. We used that figure in 
computing our conservative estimate (given that the costs associated 
with road congestion have grown since 2000) of $7,000 per million ton- 
miles shown in table 3. We found no national estimates of the external 
congestion costs waterways freight causes to passenger, recreational, 
and other nonfreight waterways users.[Footnote 57] There is no 
national policy to charge transportation infrastructure users for 
their contribution to congestion. 

Taxes and Fees Associated with Marginal Freight Activity: 

Federal, state, and local governments levy certain taxes and user fees 
on road users that increase with the payers' use. These levies include 
taxes on motor fuels and tires, as well as tolls. FHWA provided us 
with underlying data from its forthcoming highway cost allocation 
study that estimates how much of the various federal highway user 
taxes and fees are attributable to trucks. We combined this data with 
our own estimates for state and local fuel taxes and tolls in order to 
obtain our estimate of the total tax and fee payments that trucks make 
for their marginal use of highways, which amounts to about $11,000 per 
million ton-miles.[Footnote 58] In comparison, estimates by TRB and 
CBO suggest that marginal fees paid by waterways freight service 
providers are less than $500 per million ton-miles. Railroads do not 
pay taxes or fees for the marginal use of their own infrastructure. 

Consumers of Freight Services Pay Less of the Fixed Costs Associated 
with Trucking than with Railroads or Waterways: 

We examined the extent to which fixed costs are not passed on to final 
consumers separately from our table 3 comparison of marginal costs and 
marginal taxes and fees because unpriced fixed costs will not cause 
inefficient use of existing infrastructure as unpriced marginal costs 
do; however, unpriced fixed costs can lead to inefficient investment 
decisions (as discussed in the following section). Fixed public 
infrastructure costs are those, such as investments in new roads or 
the dredging of a waterway, which would exist regardless of whether an 
additional shipment is made on the route. Fixed taxes and fees, such 
as excise taxes on vehicle purchases and registration fees, do not 
vary with the number of VMT. Our estimates in table 5 indicate that 
the unpriced fixed social costs per ton-mile are largest for trucking--
$7,000 per million ton-miles--and smallest for waterways freight-- 
$2,000 per million ton-miles. Railroad infrastructure is, for the most 
part, privately owned and thus has negligible fixed public 
infrastructure costs.[Footnote 59] 

Table 5: Estimates of Fixed Social Costs Attributable to Each Freight 
Mode That Are Not Passed on to Consumers, per Million Ton-Miles: 

Monetary values (in thousands of constant 2010 dollars): 

Fixed social costs: 

Fixed private costs; 
Trucking: [A]; 
Railroad: [A]; 
Waterways: [A]. 

Fixed public infrastructure costs; 
Trucking: $14; 
Railroad: -[C]; 
Waterways: $4,000. 

Other public subsidies--federal tax subsidies and financing 
programs[B]; 
Trucking: Unknown on a per-ton-mile basis but varies little across 
modes; 
Railroad: Unknown on a per-ton-mile basis but varies little across 
modes; 
Waterways: Unknown on a per-ton-mile basis but varies little across 
modes. 

Fixed external costs; 
Trucking: Unknown, but likely small; 
Railroad: Unknown, but likely small; 
Waterways: Unknown, but likely small. 

Fixed taxes and fees: 

Taxes and fees targeted at the freight modes but not on marginal 
freight activity; 
Trucking: $7,000; 
Railroad: Unknown, but likely small; 
Waterways: $2,000. 

Fixed social costs not passed on to consumers: 

Unpriced costs--Fixed social costs minus taxes and fees (excluding 
other public subsidies); 
Trucking: $7,000; 
Railroad: -; 
Waterways: $2,000. 

Sources: GAO analysis of DOT, Department of the Treasury, Census 
Bureau, and Corps data. 

[A] As explained above, private costs are generally passed on to 
consumers, so they do not need to be added into the estimation of 
costs that are not passed on. 

[B] This category does not include any state or local government tax 
or financing subsidies. The costs of the financing programs and tax 
exempt bonds would be negligible per ton-mile. We could not estimate 
the cost of preferential treatment under the corporate income tax on a 
per-ton-mile basis; however, as explained above, this treatment varies 
little across modes. 

[C] "-" means less than .5. 

[End of table] 

Within the freight trucking mode, we also compared single-unit and 
combination trucks. Our analysis suggests that fixed costs exceeded 
the share of fixed taxes and fees for both types of trucks, and that 
the amount of these unpriced fixed costs was higher for single-unit 
trucks than for combination trucks: 7.7 cents versus 5.2 cents per 
VMT.[Footnote 60] In contrast, the marginal revenues exceeded the 
marginal costs for each type of truck, with the difference larger for 
single-unit than for combination trucks: 4.8 cents versus 3.5 cents 
per VMT. 

Formulating Policy Responses to Address Unpriced Social Costs Raises 
Complex Issues: 

In this report we have pointed out that the efficiency of our economy 
is decreased in several ways when marginal and fixed costs are not 
reflected in prices, and that the available evidence at a national 
level indicates that there are unpriced marginal and fixed social 
costs across the three surface freight transportation modes. Policy 
responses that attempt to more closely align prices with marginal 
social costs (including a competitive rate of return on capital) or 
attempt to reduce gaps between fixed costs and revenues will confront 
a number of complex issues that are important for policymakers to 
consider, particularly when considering national-level policies. We 
also noted that the extent to which modes are substitutable is 
difficult to estimate and will largely be determined by whether 
shipping is feasible or practical by another mode, and by the relative 
prices and other service characteristics of shipping by different 
modes. In addition to mode-shifting, price changes can prompt other 
economic responses in the short run, such as the use of lighter-weight 
materials; over the longer term there is greater potential for 
responses that will shape the overall distribution and use of freight 
services. 

Costs can vary widely based on the specific characteristics of an 
individual shipment, such as the geography and population density of 
the shipment's route, and the fuel-efficiency of the specific vehicle 
carrying it. Ideally, policy that is able to align marginal prices 
with marginal costs on a shipment-by-shipment basis would provide the 
greatest economic benefit. However, achieving this in practice would 
typically result in high administrative costs. For example, freight 
carriers may have to purchase new technologies or be required to 
maintain more complex and detailed records. Similarly, government 
agencies would likely have to devote more resources to enforcement 
efforts. As a result, economic efficiency could be reduced because the 
costs to administer the policy may actually exceed the benefits 
achieved. 

Less targeted interventions (e.g., charging fees or taxes based on 
average costs, subsidizing more efficient alternatives, or broadly 
applying safety or emission regulations) can have impacts on users and 
potentially change the overall distribution of freight across modes or 
demand for freight overall, but provide fewer benefits. Further, more 
general policy interventions can push too much of the cost onto users 
who previously had below-average unpriced costs and too little of the 
cost onto users who previously had above-average unpriced costs. For 
example, a policy that charges freight providers on the amount of 
their emissions would result in an overcharge for those traveling in 
rural areas where few people live and an undercharge for those 
traveling in more densely populated urban areas. External costs from 
the same amount of emissions would be higher in more densely populated 
urban areas because more people are exposed to the pollutants. 

Other complexities arise when attempting to align fixed costs and 
revenues. In general, our current system is set up as a user pay 
system, wherein the costs of building and maintaining the system are 
to be borne by those who benefit. However, available data suggest that 
in the trucking and waterways modes, current government mechanisms to 
recover the fixed costs associated with public infrastructure do not 
achieve full recovery. Aligning fixed costs and revenues for public 
infrastructure--whose multiple users include passenger cars and 
recreational boats along with freight trucks and vessels--is a complex 
task requiring detailed cost allocation studies, which are expensive 
and time-consuming, and are not done regularly.[Footnote 61] 
Furthermore, policies designed to recover fixed costs can conflict 
with policies designed to address gaps between marginal social costs 
and revenues. As discussed previously, if governments were to charge 
users only for the marginal costs of their use, in many cases they 
would not be able to recover the costs of building the infrastructure 
to begin with. However, once the infrastructure is in place, charging 
users a portion of the fixed costs each time they use the 
infrastructure (on top of a charge for any marginal costs they impose) 
would likely result in underutilization of the infrastructure because 
some potential users would not be willing to bear the higher cost. 
Appendix V provides options that governments can take to address this 
tradeoff between efficiency and cost recovery. 

Finally, marginal social costs can vary widely across jurisdictions, 
and have varying levels of impact, which has implications for the 
level of government that is best suited to administer a policy 
response. For example, congestion costs are local in nature, thus 
cities, counties, or local authorities are in the best position to 
develop interventions that reduce those costs, or attempt to price 
those costs. On the other hand, some air pollution costs can be 
imposed on multiple states, the entire nation, or other countries. 
State or local governments may not be equipped or institutionally 
capable of implementing policies that are regional in nature and 
affect multiple states. National policy responses to pollution and 
emissions must also consider that air pollution reductions can be 
achieved across a number of different industries, potentially at lower 
cost than in the transportation sector. Furthermore, although 
considerable research has gone into estimating the effects of climate 
change, there is uncertainty around how increases in atmospheric 
concentrations of greenhouse gases and temperature within ecosystems 
and economic growth will vary across regions, countries, and economic 
sectors, and therefore, appropriate policy responses require 
international coordination.[Footnote 62] 

Agency Comments: 

We provided copies of a draft of this report to DOT and the Corps for 
review and comment. DOT responded with suggestions to consider 
additional data sources and methods for calculating infrastructure and 
external costs. We accepted some of DOT's suggestions and incorporated 
those changes into our report, but for others, we believe that our 
data sources and methods are appropriate. DOT also provided technical 
corrections, which we incorporated in the report. The Corps indicated 
that we had adequately incorporated or footnoted its comments made to 
a preliminary draft of this report and had no further comment. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies to the 
Secretary of Transportation, the Secretary of Defense, the Commanding 
General and Chief of Engineers of the U.S. Army Corps of Engineers, 
and interested congressional committees. In addition, the report will 
also be available at no charge on the GAO Web site at [hyperlink, 
http://www.gao.gov]. 

If you have any questions about this report, please contact our 
offices at (202) 512-2843 or (202) 512-9110 or at herrp@gao.gov or 
whitej@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 VI. 

Signed by: 

Phillip R. Herr: 
Director, Physical Infrastructure: 

Signed by: 

James R. White: 
Director, Tax Issues: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

The objectives of this report are to (1) describe how public policies 
can affect competition and efficiency within the surface freight 
transportation sector; (2) determine what is known about the extent to 
which costs are borne by surface freight users; and (3) discuss how 
our findings could be used when making future surface freight 
transportation policy. 

To describe the effects of public policy on the freight sector, we 
conducted a review of the transportation and economic literature and 
interviewed transportation policy experts to identify how government, 
academic, and professional research organizations apply economic 
concepts to determine the efficiency of the surface freight 
transportation system in the United States. We identified the types of 
data that can be used to evaluate the costs imposed by users of the 
surface freight transportation system on the economy and the factors 
to consider when determining the effect of government intervention. 

To determine the extent to which costs are borne by users of freight 
trucking, freight rail, and waterways freight services, we obtained, 
reviewed, and analyzed several datasets. We used federal highway cost 
and revenue data provided by the Federal Highway Administration's 
(FHWA) Office of Policy for allocating costs imposed by and estimating 
revenue received from the freight trucking industry--specifically, 
single-unit and combination trucks. For allocating similar costs and 
revenues to the same classes of trucks at the state and local 
government levels, we used several tables reported by FHWA's Highway 
Statistics Series. To estimate domestic waterways costs at the federal 
level, we used construction costs reported by the U.S. Army Corps of 
Engineers (Corps), and used operations and maintenance costs obtained 
from the Corps' Operations and Maintenance Business Information Link 
system. For estimating state and local governments' financial 
assistance to waterways, we used the Census Bureau's State and Local 
Government Finance data. In terms of revenue, inland waterways and the 
harbor maintenance trust fund revenue estimates were based on 
financial information reported by the Department of the Treasury's 
Bureau of Public Debt. We used available cost and revenue information 
on railroads published by the Association of American Railroads. To 
assess the reliability of finance and technical data, largely gathered 
from federal statistical agencies' databases, we reviewed relevant 
documentation about the agencies' data collection and quality 
assurance processes, talked with knowledgeable officials from several 
agencies about these data, and compared these data against other 
sources of published information to determine data consistency and 
reasonableness. We determined that the data were sufficiently reliable 
for the purposes of this report. 

We reviewed marginal and fixed costs from the literature, including 
(1) the 1997 FHWA Highway Cost Allocation Study, its addendum 
published in 2000, and unpublished supporting documentation; (2) the 
Transportation Research Board's 2008 National Cooperative Highway 
Research Program, Synthesis 378, on State Highway Cost Allocation 
Studies; (3) the 1996 Transportation Research Board (TRB) report 
entitled Paying Our Way, Estimating Marginal Social Costs of Freight 
Transportation; (4) the 1992 Congressional Budget Office (CBO) report 
entitled Paying for Highways, Airways, and Waterways: How Can Users Be 
Charged; (5) the American Association of State Highway and 
Transportation Officials publications; and (6) Road Work, A New 
Highway Pricing and Investment Policy, published by the Brookings 
Institution in 1989. 

To understand available financial and technical data on freight, we 
conducted interviews with and obtained data from officials in the 
following organizations: 

* Department of Transportation's Chief Economist, Federal Highway 
Administration's Freight Management of Operations, and Transportation 
Infrastructure and Finance Innovation Act program office; 

* Federal Railroad Administration's Railroad Rehabilitation and 
Improvement Financing program office; 

* Research and Innovation Technology Administration's Bureau of 
Transportation Statistics; 

* Maritime Administration's ship financing program office; 

* Army Corps of Engineers' Institute for Water Resources; and: 

* Environmental Protection Agency's (EPA) Office of Air and Radiation. 

We also interviewed Department of Transportation (DOT), Corps, and EPA 
officials to obtain advice on economic concepts related to surface 
freight transportation, appropriate and available data sources, and 
methodological approaches. We obtained preliminary reviews about the 
scope, methodology, and analysis contained in this report from DOT, 
EPA, the Corps, as well as two members of the Comptroller General's 
Advisory Board--comprised of individuals with broad expertise in 
public policy. 

We also spoke with industry representatives to discuss their views on 
government spending and regulatory policy. Specifically, we 
interviewed representatives from the American Trucking Association, 
the Association of American Railroads, the American Waterways 
Operators, and the American Association of Port Authorities. In 
addition to industry representatives, we also interviewed individuals 
who were involved in previous federal and state highway cost 
allocation studies or authored research papers on external costs. 

Cost Estimation Methodology: 

We estimated federal, state, and local government costs for the 
surface freight transportation infrastructure, including the publicly 
owned highways and domestic, commercial inland waterways by examining 
government-reported spending data. Freight railway infrastructure, on 
the other hand, is, for the most part, privately owned and operated. 
Private railroad investment costs and revenues are proprietary, and 
therefore, we did not attempt to produce estimates of private costs 
and limited our analysis to government expenditures associated with 
rail, where appropriate. Further, we did not consider in our study (1) 
pipeline freight because pipelines carry specific liquid commodities, 
such as natural gas and oil products or (2) air freight because air 
freight constitutes a fraction of commercial freight moved by value, 
ton, and ton-miles and is typically used to move high-value, time- 
sensitive freight which would generally not be moved by the other 
modes. Where possible, we adjusted all figures to constant 2010 
dollars using the fiscal year gross domestic product price index. 

In compiling our results for freight trucking, we could only 
approximate the freight truck population by using available data for 
all single-unit and all combination trucks--any vehicle consisting of 
a power unit pulling at least one trailer that does not have a power 
unit--but excluded light trucks, which are generally passenger 
vehicles and delivery vans. In using this population to represent 
freight trucks, we include some trucks that are involved in nonfreight 
purposes, such as municipal waste disposal trucks and utility trucks, 
which account for a small percentage of the total vehicle miles 
traveled (VMT) of this population.[Footnote 63] Some of the nonfreight 
trucks are likely to have marginal costs and tax payments that are 
lower than those for the average freight truck, while other nonfreight 
trucks will have higher costs and tax payments. While the population 
of freight trucks could also have been defined based on the number of 
axles on a truck, FHWA's 1997 Highway Cost Allocation Study--the basis 
for some of our spending and revenue projections--reported costs and 
revenues by various weights of single-unit and combination trucks. 

Our estimates of external costs--the costs imposed on society, such as 
the cost of lost time resulting from traffic congestion or the health 
consequences related to pollution--reflect activities that can be 
attributed to domestic freight activity for all single-unit and 
combination trucks, rail carriers, and waterborne vessels. In several 
instances, we made adjustments to national data in an effort to remove 
nondomestic or nonfreight activity from our calculations. For example, 
as shown in table 9 in this appendix, we adjusted EPA data to more 
accurately report emissions attributable to domestic freight activity 
for all three modes. 

Both our estimates of government costs and revenues and external costs 
are based on high-level data in order to compare the modes on a 
nationwide basis. Variations in costs and revenue across individual 
shipments within each mode may be obscured by this level of 
aggregation. Because these are comparisons between modes on an 
aggregate, national basis, and we are not able to compare specific 
shipments, the estimates associated with railroads and waterways do 
not consider the costs associated with the truck ton-miles necessary 
to complete a shipment on those modes (for an analysis that attempts 
to compare marginal costs across the modes on a shipment-by-shipment 
basis, see Transportation Research Board, Paying Our Way, Estimating 
Marginal Social Costs of Freight Transportation (1996)). The results 
should be viewed as representing averages across all of the marginal 
shipments that were made under a wide variety of different conditions 
in a wide variety of locations. 

Federal Highway Marginal and Fixed Costs and Revenues: 

FHWA has conducted highway cost allocation studies--the most recent 
being in 1997, which superseded its 1982 study--that attempted to 
determine whether all highway users are paying their fair share of 
federal highway costs and to ensure that it and Congress have up-to- 
date information when making future decisions affecting federal 
highway user fees. According to FHWA officials, sections of the 1997 
report were peer-reviewed by TRB, and based on TRB's comments, FHWA 
issued an addendum in 2000. According to FHWA officials, an update to 
this study is forthcoming. To the extent possible, we developed cost 
and revenue categories similar to those used in FHWA's 1997 study. 

For this review, we obtained FHWA's Office of Policy data on average 
spending from 2000 through 2006 on highways by improvement type and 
vehicle class--specifically single-unit and combination trucks. 
Improvement types included new construction, preservation, minor 
widening, bridge work, safety and traffic operations, and 
environmental, among others. We also obtained federal revenue data by 
vehicle class and revenue type--fuel, retail, heavy vehicle use, and 
tire taxes. We used this FHWA data to estimate total federal highway 
costs and revenues attributed to freight trucks. 

We separated the federal costs and revenues into two categories--the 
first included costs and revenues associated with marginal use of the 
highways by freight trucks. We considered highway system preservation 
costs to be the closest available approximation of the wear-and-tear 
costs associated with road use.[Footnote 64] However, we did not 
include (1) bridge-related costs because bridges are built to 
withstand a specific design load[Footnote 65] or (2) enhancement or 
new capacity costs because these costs do not directly vary with 
repeated truck usage. All other federal spending was considered to be 
fixed costs. Our use of spending data to represent marginal costs may 
result in an understatement of those costs if that spending was not 
sufficient to repair all of the damage caused by road use. We 
considered marginal revenue to be receipts from fuel and tire taxes--
receipts directly related to the use of the highway infrastructure. We 
assumed the retail tax had little relationship to highway use, and 
therefore, did not include it in our marginal revenue category. 

State Highway Marginal and Fixed Costs and Revenues: 

Because FHWA efforts to update the 1997 study will not address state 
and local government costs, we produced our own estimates of state and 
local government costs and revenues allocated to freight trucks for 
the same time period as the federal-level data that FHWA had provided 
to us. For our state estimates, we summarized state expenditures on 
highways from fiscal years 2000 through 2006 using table SF-12A, State 
Highway Agency Capital Outlay, from FHWA's Highway Statistics Series. 
With assistance from FHWA officials, we categorized these costs into 
four improvement types--new capacity, system preservation, 
enhancements, and other--consistent with cost categories identified in 
the 1997 Highway Cost Allocation Study and in federal costs reported 
in the Highway Statistics Series. Because state expenditures reported 
in table SF-12A included federal funds that states received, we 
adjusted the expenditures to reflect strictly state spending on 
highways. First, we converted federal obligations reported in highway 
statistics table FA-6A (Obligation of Federal Funds) to expenditures 
using factors provided by an official from FHWA.[Footnote 66] Second, 
for each of the four improvement types, we subtracted the estimated 
federal expenditures from state expenditures to obtain states' 
spending of their own funds. Third, for each improvement type, we 
estimated the expenditure amounts attributable to freight trucks using 
proportions that we derived from supporting documentation related to 
the 1997 study. State and local governments are generally responsible 
for maintaining the nation's highways, and therefore, we again used a 
proportion derived from data from the 1997 study and applied it to 
Highway Statistics Series table SF-2 (State Disbursements for 
Highways) containing maintenance and services figures to estimate 
operations and maintenance costs attributed to freight trucking at the 
state level.[Footnote 67] As with the federal spending data, we 
considered system preservation expenditures to be the best available 
approximation of costs associated with marginal highway use. We 
consider all other costs, such as new construction, system 
enhancements, and routine maintenance to be fixed costs.[Footnote 68] 

To estimate marginal state revenues attributable to freight trucks 
from fiscal years 2000 through 2006, we first determined the average, 
annual total receipts from motor fuels receipts (minus penalties and 
fines) reported in table MF-1 (State Motor Fuel Taxes and Related 
Receipts) and tolls from bridge, tunnel, and road crossings receipts 
reported in table SF-3B (State Administered Toll Road and Crossing 
Facilities). We then determined what shares of these revenues were 
attributed to freight trucks as follows: except for tolls, the revenue 
shares from fuel for freight trucks were based on results from the 
1997 Highway Cost Allocation Study.[Footnote 69] The revenue shares 
for each year after 2000 were adjusted for changes in VMT, motor fuels 
consumed, and vehicle registrations. Since the 1997 study did not 
allocate tolls, we assumed that for each freight truck category, the 
share of tolls was equal to its share of total VMT in a given year. 
For fixed revenues, we also included registration, drivers license, 
and weight-distance receipts, as reported in table MV-2 (State Motor-
Vehicle and Motor-Carrier Tax Receipts), in our calculations. 

Local Highway Infrastructure Marginal and Fixed Costs and Revenues: 

We also developed marginal and fixed cost estimates for local highway 
spending and revenues attributed to freight trucks. For local 
expenditures, we summarized local disbursements on highways, averaged 
from 2000 through 2006 using table LGF-2 from the Highway Statistics 
Series. The table grouped data by capital outlays, maintenance and 
traffic services, administration and miscellaneous, and law 
enforcement, among other categories. We grouped the data as closely as 
we could into categories approximating those that FHWA used in their 
1997 study and then allocated these disbursements to single-unit and 
combination trucks based on those trucks' shares of the 1997 
categories. Given that capital construction and system preservation 
was reported as a single category, the only option we had for 
estimating the amount spent on pavement rehabilitation (which we used 
to benchmark marginal costs) was to assume that pavement 
rehabilitation accounted for the same share of total capital costs as 
it did in 2000. On the revenue side, we again used tables from the 
Highway Statistics Series: table LDF (Local Government Receipts from 
State and Local Highway User Revenue) and table LGF-3B (Receipts of 
Local Toll Facilities). Table LDF reported motor fuel and motor 
vehicle revenues combined as a single number. Using data from the 1997 
cost study, we estimated that motor fuel accounted for 45 percent of 
combined motor fuel and motor vehicle revenue in 1994. We used the 
trucks' shares of local motor fuel and motor vehicle taxes from the 
1997 study to allocate our updated revenue amounts to trucks. For 
tolls, we used the same allocation assumption as we did for state toll 
revenues. Local revenues account for less than 3 percent of the 
marginal revenue and fixed revenue amounts that we report in tables 3 
and 5 respectively. Table 6 summarizes expenditures and revenues by 
mode by government per million ton-miles. 

Table 6: Estimated Average Infrastructure Expenditures and Revenue 
(per Million Ton-Miles) by Level of Government by Mode: 

Monetary values (in thousand of constant 2010 dollars): 

Costs: 

Mode: Trucks--related to marginal costs; 
Federal: $3,000; 
State: $2,000; 
Local: $1,000; 
All levels of government[A]: $7,000. 

Mode: Trucks--related to fixed costs; 
Federal: $4,000; 
State: $4,000; 
Local: $5,000; 
All levels of government[A]: $14,000. 

Mode: Railroads; 
Federal: Not estimated; 
freight rails receive limited government assistance; 
State: Not estimated; 
freight rails receive limited government assistance; 
Local: Not estimated; 
freight rails receive limited government assistance; 
All levels of government[A]: Not estimated; 
freight rails receive limited government assistance. 

Mode: Waterways; 
Federal: $2,000; 
State: $3,000 (State and local); 
Local: [Empty]; 
All levels of government[A]: $4,000. 

Revenues: 

Mode: Trucks--related to marginal revenues; 
Federal: $6,000; 
State: $5,000; 
Local: -[B]; 
All levels of government[A]: $11,000. 

Mode: Trucks--related to fixed revenues; 
Federal: $2,000; 
State: $5,000; 
Local: -[B]; 
All levels of government[A]: $7,000. 

Mode: Railroads; 
Federal: Not estimated; 
freight rails pay some taxes; 
State: Not estimated; 
freight rails pay some taxes; 
Local: Not estimated; 
freight rails pay some taxes; 
All levels of government[A]: Not estimated; 
freight rails pay some taxes. 

Mode: Waterways; 
Federal: -[B]; 
State: $2,000 (State and local); 
Local: [Empty]; 
All levels of government[A]: $2,000. 

Sources: GAO calculations based on DOT, Corps, and Census data. 

[A] Federal, state, and local expenditures may not total the all level 
of government expenditures due to rounding. 

[B] "-" means less than .5. 

[End of table] 

Freight Rail Costs and Revenues: 

All Class I railroad infrastructure is privately owned, and most other 
classes of railroads are also privately owned. However, the federal 
government provides some limited assistance to privately owned 
railroads, but it was negligible for the purposes of our analysis. The 
federal government no longer levies any federal excise tax on 
railroads.[Footnote 70] There is little available evidence on the 
extent to which railroads receive financial assistance from states or 
local governments, but this evidence suggests that these amounts are 
negligible. Railroads are subject to state and local property taxes on 
their infrastructure and the nation's major railroads paid at least 
$625 million in 2008. 

Waterways Freight Costs: 

The nation's waterways are used for many purposes, such as navigation, 
flood control, irrigation, and recreation. According to literature we 
reviewed, the marginal infrastructure costs associated with freight on 
the waterways are negligible.[Footnote 71] To estimate the overall 
fixed cost to the federal government for waterway infrastructure 
investments, as well as operations and maintenance in support of 
freight transportation, we obtained budgeting and expenditure data by 
waterway (deep and shallow draft coastal harbors and channels, Great 
Lakes, and inland waterways) from fiscal years 2000 through 2006 from 
the Corps and the Saint Lawrence Seaway Development Corporation. 

* Coastal harbors and channels operations and maintenance, 
investments, and nonfuel taxed waterways investments. We allocated the 
Corps' total operations and maintenance harbors and channels (deep and 
shallow draft) expenditures by year to each state based on the Harbor 
Maintenance Trust Fund expenditures by state.[Footnote 72] We also 
obtained waterway infrastructure investment costs by project by state 
from the Corps. We allocated the expenditures to domestic freight 
based on tonnage--specifically, the percent of total tonnage moved 
through each state that is domestic waterways freight--as reported by 
Corps data supporting table 4-1, Waterborne Commerce by States, 
Waterborne Commerce of the United States, National Summaries. Such 
allocation may overstate costs attributable to domestic freight 
operations for at least two reasons. First, CRS reports that a 
significant amount of harbor spending is directed toward harbors that 
handle little cargo, and therefore the primary beneficiaries of the 
spending will be nonfreight users of the harbors and channels. Second, 
dredging at U.S. ports may be done primarily to accommodate ever-
larger container ships involved in oceanic trade, and therefore costs 
attributable to domestic trade may be negligible in those cases. 
However, without a waterways cost allocation study, little more is 
known about how costs may be distributed among the various users of 
harbors, channels, and other waterways, and thus tonnage appears to be 
the most reasonable method to allocate costs. 

* Inland waterways operations, maintenance, and construction costs. 
After consultations with a Corps official, we allocated 50 percent of 
the Corps' inland waterways operations, maintenance, and construction 
spending from fiscal years 2000 through 2006 to freight. We used 50 
percent because federal law establishes a 50/50 federal/nonfederal 
cost-share arrangement for construction. We could not definitively 
determine the extent to which waterways freight activity accounted for 
all waterway activity. Two studies provide a wide range of 
possibilities. A 1980 Corps study indicated that for selected 
waterways (the Ohio, Allegheny, Monongahela, Lower and Upper 
Mississippi, among others) within the boundaries of three Corps 
districts--St. Paul, St. Louis, and Pittsburgh--the average waterways 
freight activity accounted for 75 percent.[Footnote 73] More recently, 
however, a 2010 preliminary report concluded that a wide range of 
consumers benefit from the pools of water created and operated to 
facilitate commercial navigation and other uses, but commercial 
navigation itself appears to be a relatively small beneficiary of this 
system. This finding was based on a limited scope of work, and without 
further research, allocating costs or revenues to commercial freight 
has limitations. 

State and local governments spend their own funds for investments in 
state-owned port facilities involved in domestic freight 
transportation. For our state and local government analysis, we used 
expenditure and revenue data on "sea and inland port facilities" from 
the U.S. Census Bureau's, Government Finance Statistics, State and 
Local Government Finances by Level of Government. States and local 
governments provide funding to publicly owned ports and dock 
facilities on waterways for the purposes of construction, operation, 
and maintenance of commercial port facilities, canals, harbors, and 
other public waterways; dredging of those waterways; and maintenance 
of commercial docks, piers, wharves, warehouses, cranes, and 
associated terminal facilities, among other things. To determine the 
portion of this spending that may be allocated to domestic waterways 
freight transportation, we used factors based on tonnage--
specifically, the percent of total tonnage moved through the port that 
is domestic waterways freight, as reported by Corps data supporting 
table 4-1, Waterborne Commerce by States, Waterborne Commerce of the 
United States, National Summaries. 

Ton-Mile Adjustments: 

We used ton-miles to normalize our data across modes. Multiple ton-
mile estimates are available for domestic freight activity. To the 
extent possible, we attempted to use ton-mile data that most 
accurately reflects the total domestic freight activity within each 
mode. 

* Freight trucks. We used truck ton-mile estimates based on DOT's 
Freight Analysis Framework (FAF).[Footnote 74] According to DOT 
officials, the 2007 ton-mile estimate derived from the FAF are the 
most comprehensive representation of domestic truck freight activity 
available. DOT has another series of ton-mile estimates produced by 
the Bureau of Transportation Statistics (BTS); however, according to 
DOT officials the BTS series does not capture as much domestic truck 
freight activity as the FAF estimate.[Footnote 75] We determined that 
the FAF data were more appropriate for the purpose of presenting our 
cost and revenue data on a per-ton-mile basis because the cost and 
revenues data we used were for the broadest definition of truck 
freight traffic.[Footnote 76] One difficulty in using the FAF 
estimates is that 2007 is the only recent year for which DOT has 
applied the current FAF methodology. DOT in previous years applied a 
different methodology to estimate ton-miles based on 2002 data. 
However, given that the methodology for estimating these figures 
changed significantly between 2002 and 2007, DOT cautions that the 
estimates from the 2 years should not be combined in the same time 
series.[Footnote 77] In order to produce ton-mile estimates for all of 
the years that we needed, we multiplied the BTS figure for each year 
by the ratio of the FAF estimate to the BTS estimate for 2007. Given 
the unavoidable imprecision of this approach, we report error bounds 
of plus and minus 5 percent for all of our per-ton-mile results. These 
ton-mile data are shown in table 7. 

Table 7: Estimated Truck Ton-Miles of Domestic Surface Freight, 2000 - 
2007: 

Ton miles (in millions): 

Ton-mile source: Freight Analysis Framework (FAF); 
2000: [Empty]; 
2001: [Empty]; 
2002: [Empty]; 
2003: [Empty]; 
2004: [Empty]; 
2005: [Empty]; 
2006: [Empty]; 
2007: 2,040,000. 

Ton-mile source: GAO estimate based on 2007 FAF/BTS ratio (1.55); 
2000: 1,847,273; 
2001: 1,878,839; 
2002: 1,928,914; 
2003: 1,958,696; 
2004: 1,984,713; 
2005: 2,000,110; 
2006: 2,000,011; 
2007: [Empty]. 

Ton-mile source: Plus 5 percent; 
2000: 1,939,636; 
2001: 1,972,781; 
2002: 2,025,359; 
2003: 2,056,631; 
2004: 2,083,949; 
2005: 2,100,115; 
2006: 2,100,012; 
2007: [Empty]. 

Ton-mile source: Minus 5 percent; 
2000: 1,754,909; 
2001: 1,784,897; 
2002: 1,832,468; 
2003: 1,860,762; 
2004: 1,885,477; 
2005: 1,900,104; 
2006: 1,900,011; 
2007: [Empty]. 

Ton-mile source: Bureau of Transportation Statistics (BTS); 
2000: 1,192,633; 
2001: 1,213,013; 
2002: 1,245,342; 
2003: 1,264,570; 
2004: 1,281,367; 
2005: 1,291,308; 
2006: 1,291,244; 
2007: 1,317,061. 

Source: GAO analysis of DOT data. 

[End of table] 

* Freight rail. We used freight ton-miles reported in table 1-46b from 
BTS's 2009 National Transportation Statistics report. 

* Waterways freight. We used ton-miles for all domestic waterways, 
including the inland waterways (internal or intraport), coastal 
waterways (coastwise), and Great Lakes (lakewise), as reported in 
table 1-4 of the Corps' 2008 Waterborne Commerce of the United States, 
National Summary. When estimating the marginal external costs for 
particulate matter and nitrogen oxide, we strictly used the ton-miles 
along the inland waterways because the available data included only 
the pollution along the inland waterways system. 

External Costs: 

To determine what is known about freight external costs and how 
government policies shift costs to freight users, we analyzed and 
synthesized cost estimates reported for each transportation mode and 
calculated accident and pollution incident rates. We reviewed reports 
and studies issued by federal agencies, transportation research 
organizations, and academia, as well as our past work in surface and 
freight transportation and the environment. We also discussed freight 
transportation externalities and policies with a number of 
knowledgeable government and non-government officials. 

To describe freight accident external costs among modes, we reviewed 
the available estimates. We found significant variation across study 
methodologies, such as what segments of freight transportation were 
included, whether freight operator accidents or injuries were 
included, and whether they estimated average or marginal costs. 
Additionally, all of the estimates were dated (based on freight 
activities from 2000 or earlier). To determine whether recent rates 
were consistent with previous cost estimates. We calculated more 
recent national rates of accident fatalities and injuries involving a 
freight carrier, in ton-mile terms, for each mode during calendar 
years 2003 to 2007. Table 8 depicts our approach to calculating these 
rates. 

Table 8: Methodology for Estimating Average Annual Accident Fatalities 
and Injuries, per Billion Ton-Miles, Average of 2003 to 2007: 

Mode: Trucks[A]; 
Fatalities: 5,069; 
Injuries: 111,800; 
Estimated billion ton-miles: 1,997; 
Fatalities per billion ton-miles: 2.54; 
Injuries per billion ton-miles: 56.05. 

Mode: Trains[B]; 
Fatalities: 683; 
Injuries: 5,747; 
Estimated billion ton-miles: 1,739; 
Fatalities per billion ton-miles: 0.39; 
Injuries per billion ton-miles: 3.32. 

Mode: Waterborne vessels[C]; 
Fatalities: 7; 
Injuries: 26; 
Estimated billion ton-miles: 587; 
Fatalities per billion ton-miles: 0.01; 
Injuries per billion ton-miles: 0.05. 

Source: GAO analysis of DOT 2010 data. 

[A] Fatalities and injuries reported in Federal Motor Carrier Safety 
Administration's Federal Motor Carrier Safety Administration Large 
Truck and Bus Crash Facts 2007 (table 1 and table 4). Trucks are 
defined as over 10,000 gross vehicle weight, which can include some 
nonfreight activity. For example, in 2007, 12.3 percent of large 
trucks involved in a fatal accident and 13.2 percent involved in 
accidents with injuries were dump, garbage or concrete mixer trucks. 

[B] Fatalities and injuries reported in Federal Railroad 
Administration Office of Safety Analysis's accident/incident online 
data reporting system, table 1.07: 

[C] Fatalities and injuries reported in FHWA's Freight Facts and 
Figures, tables 5-1 and 5-2. We selected freight vessels listed as 
tows, tugs, or barges to use in our analysis. 

[End of table] 

We report these computations as the nonmonetized indicators of 
relative freight external accident costs among the modes. They can be 
considered reliable indicators of external costs, since these are 
accident consequences that result despite regulation and other safety 
measures. Available estimates used different methods and assumptions 
for determining what portion of total accident costs is external, and 
estimates varied from 48 to 62 percent of total accident costs. We 
concluded that none of the available evidence about the external costs 
portion would significantly change the disparity in accident costs 
between truck freight and the other two modes that is depicted by the 
overall accident rates. 

To describe freight pollution external costs among modes, we reviewed 
the existing literature and estimates. Other studies estimated average 
external costs for intercity truck and rail freight, and found that 
intercity truck freight could be as high as 1.67 cents per ton-mile 
and rail freight could be as high as .38 cents per ton-mile in 
constant 2010 dollars. Other related publications variously report 
emissions information for one or more of the three modes, but not 
economic costs. 

Table 9 depicts our approach to estimating national rates of emissions 
for two key regulated emissions typically comprising the majority of 
estimated air pollution external costs--nitrogen oxide (NOX) and fine 
particulate matter with a diameter of 2.5 microns or less (PM2.5)--in 
ton-mile terms for each mode. 

Table 9: Methodology for Estimating Tons of Freight-Related PM2.5 and 
NOX Emissions, per Million Freight Ton-Miles for Trucks and 
Locomotives in 2002 and for Waterborne Vessels in 2005: 

Mode: Trucks[A]; 
Estimated tons of PM2.5 emissions: 229,754; 
Estimated tons of NOX emissions: 5,824,060; 
Estimated millions of ton-miles: 1,928,914; 
Estimated tons of PM2.5 per million ton-miles: 0.1191; 
Estimated tons of NOX per million ton-miles: 3.0193. 

Mode: 5 percent ton-mile increase; 
Estimated tons of PM2.5 emissions: 229,754; 
Estimated tons of NOX emissions: 5,824,060; 
Estimated millions of ton-miles: 2,025,359; 
Estimated tons of PM2.5 per million ton-miles: 0.1134; 
Estimated tons of NOX per million ton-miles: 2.8756. 

Mode: 5 percent ton-mile decrease; 
Estimated tons of PM2.5 emissions: 229,754; 
Estimated tons of NOX emissions: 5,824,060; 
Estimated millions of ton-miles: 1,832,468; 
Estimated tons of PM2.5 per million ton-miles: 0.1254; 
Estimated tons of NOX per million ton-miles: 3.1783. 

Mode: Rail locomotives[B]; 
Estimated tons of PM2.5 emissions: 28,690; 
Estimated tons of NOX emissions: 1,083,320; 
Estimated millions of ton-miles: 1,605,532; 
Estimated tons of PM2.5 per million ton-miles: 0.0179; 
Estimated tons of NOX per million ton-miles: 0.6747. 

Mode: Waterborne vessels[C]; 
Estimated tons of PM2.5 emissions: 3,520; 
Estimated tons of NOX emissions: 141,865; 
Estimated millions of ton-miles: 274,367; 
Estimated tons of PM2.5 per million ton-miles: 0.0116; 
Estimated tons of NOX per million ton-miles: 0.4691. 

Source: GAO analysis of EPA and Texas Transportation Institute data. 

[A] Estimated emissions data are obtained directly from EPA and are 
based on the current MOVES2010 model for estimating on-road vehicle 
emissions. The estimate assumes that nearly all on-road diesel 
emissions are freight-related and 15 percent of gasoline powered 
vehicle emissions are freight-related. 

[B] Estimated emissions data are derived by subtracting 3.17 percent 
from total emissions from the locomotive category in EPA table 3-95 
and table 3-96 of EPA's Assessment and Standards Division Office of 
Transportation and Air Quality, Regulatory Impact Analysis: Control of 
Emissions of Air Pollution from Category 3 Marine Diesel Engines, 
Chapter 3 Emission Inventory, EPA-420-R-09-019. According to EPA 
documentation we reviewed 3.17 percent of total estimated locomotive 
emissions can be attributed to nonfreight activities. 

[C] The estimate is for inland waterways freight only because of 
insufficient data for other domestic waterways freight. Emissions data 
for waterways freight are for 2005 and were obtained from the Texas 
Transportation Institute, A Modal Comparison of Domestic Freight 
Transportation Effects on the General Public (2009). 

[End of table] 

Table 10 depicts our approach to monetizing the negative health 
effects associated with NOX and PM2.5 surface freight emissions. To 
monetize these effects for each freight mode, we used EPA's estimate 
of the benefits-per-ton of reducing NOX and PM2.5 emissions for 2015. 
Though EPA's benefits-per-ton estimates were intended to value 
improvements in air quality associated with emissions reductions, we 
believe that they also serve as a close approximation of the monetized 
impact of emissions increases that occur on the margin. The monetized 
value of emissions increases is referred to here as estimated 
damages.[Footnote 78] To make our calculations, we adjusted these 
estimates for both NOX and PM2.5 to 2010 dollars and then multiplied 
them by the total amount of emissions for each mode for 2002. Then, we 
divided this total by the estimated total number of ton-miles for each 
mode in 2002. (Total emissions and ton-miles are both reported in 
table 9). 

Table 10: Methodology for Estimating Damages of Freight-Related PM2.5 
and NOX Emissions, per 2002 Ton-Miles for Trucks and Locomotives and 
2005 Ton-Miles for Waterborne Vessels: 

Monetary values (in constant 2010 dollars)[A]. 

Trucks; 
Estimated damages per ton of NOX emissions: $4,610; 
Estimated damages per ton of PM2.5 emissions: $251,466; 
Estimated damages from NOX per million ton-miles: $13,920; 
Estimated damages from PM2.5 per million ton-miles: $29,950; 
Total estimated damages from NOX and PM2.5 per million ton-miles: 
$43,870. 

5 percent ton-mile increase; 
Estimated damages per ton of NOX emissions: $4,610; 
Estimated damages per ton of PM2.5 emissions: $251,466; 
Estimated damages from NOX per million ton-miles: $13,260; 
Estimated damages from PM2.5 per million ton-miles: $28,530; 
Total estimated damages from NOX and PM2.5 per million ton-miles: 
$41,780. 

5 percent ton-mile decrease; 
Estimated damages per ton of NOX emissions: $4,610; 
Estimated damages per ton of PM2.5 emissions: $251,466; 
Estimated damages from NOX per million ton-miles: $14,650; 
Estimated damages from PM2.5 per million ton-miles: $31,530; 
Total estimated damages from NOX and PM2.5 per million ton-miles: 
$46,180. 

Locomotives; 
Estimated damages per ton of NOX emissions: $4,610; 
Estimated damages per ton of PM2.5 emissions: $251,466; 
Estimated damages from NOX per million ton-miles: $3,110; 
Estimated damages from PM2.5 per million ton-miles: $4,490; 
Total estimated damages from NOX and PM2.5 per million ton-miles: 
$7,600. 

Waterborne vessels; 
Estimated damages per ton of NOX emissions: $4,610; 
Estimated damages per ton of PM2.5 emissions: $251,466; 
Estimated damages from NOX per million ton-miles: $2,380; 
Estimated damages from PM2.5 per million ton-miles: $3,230; 
Total estimated damages from NOX and PM2.5 per million ton-miles: 
$5,610. 

Source: GAO calculations based on DOT and EPA data. 

[A] We adjusted to constant 2010 dollars the benefit estimates (in 
2007 dollars) for each ton reduction in NOX and PM2.5 emissions for 
2015 (with a discount factor of 7 percent) from EPA, Final Rulemaking 
to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards and 
Corporate Average Fuel Economy Standards Regulatory Impact Analysis, 
EPA-420-R-10-009 (April 2010). 

[End of table] 

Manmade greenhouse gases include CO2, methane, nitrous oxide and 
fluorinated gases, and the dominant greenhouse gas emission for the 
transport sector is CO2. We calculated tons of CO2 equivalent 
emissions per million ton-miles of transported freight for 2007. Table 
11 depicts our approach to calculating these rates. 

Table 11: Methodology for Estimating Freight-Related CO2 Emissions, 
per 2007 Ton-Miles for Trucks and Locomotives and 2005 Ton-Miles for 
Waterborne Vessels: 

Mode: Trucks; 
Estimated tons of emissions: 468,702,769; 
Estimated ton-miles (in millions): 2,040,000; 
Estimated tons of emissions per million ton-miles: 229.76. 

Mode: Locomotives; 
Estimated tons of emissions: 52,690,481; 
Estimated ton-miles (in millions): 1,819,633; 
Estimated tons of emissions per million ton-miles: 28.96. 

Mode: Waterborne vessels[A]; 
Estimated tons of emissions: 5,286,614; 
Estimated ton-miles (in millions): 274,367; 
Estimated tons of emissions per million ton-miles: 19.27. 

Source: GAO calculations based on EPA, Inventory of U.S. Greenhouse 
Gas Emissions and Sinks: 1990-2007, table A-101 and Texas 
Transportation Institute data. 

[A] Estimate is for inland waterways freight only because sufficient 
data were not available for other domestic waterways freight. 
Emissions data for waterways freight are for 2005 and were obtained 
from the Texas Transportation Institute, A Modal Comparison of 
Domestic Freight Transportation Effects on the General Public (2009). 

[End of table] 

We conducted our review from August 2009 to January 2011 in accordance 
with generally accepted government auditing standards. Those 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: Recovery Act Funds for Freight Transportation 
Infrastructure: 

The American Recovery and Reinvestment Act of 2009 (Recovery Act) 
[Footnote 79] provided one-time funding to promote job preservation 
and creation and infrastructure investments, among other things. Since 
a portion of these funds were targeted for transportation 
infrastructure projects, generally, they may benefit both passenger 
and freight users. For example, where freight and passenger trains 
share tracks, the High-Speed Intercity Passenger Rail program may also 
enhance capacity for freight rail lines. In this appendix, we report 
funds identified for infrastructure projects and do not attempt to 
identify funding to freight or nonfreight users of the infrastructure. 
We also report Recovery Act funds identified for the EPA's Clean 
Diesel Program which helps reduce emissions for freight vehicles 
across all modes and for Army Corps of Engineers waterway projects 
that we identified as pertaining to freight transportation. 

In addition, the Build America Bonds program, created by the Recovery 
Act,[Footnote 80] allowed state and local governments to obtain 
financing at lower borrowing costs for new capital projects such as 
the development and construction of transportation infrastructure by 
having the Department of the Treasury make a direct payment to the 
state or local governmental issuer in an amount equal to 35 percent of 
the interest payment on the bonds. We summarize this information in 
table 12. 

Table 12: Summary of Recovery Act Funds for Transportation 
Infrastructure Projects That Might Benefit Specific Modes: 

Trucking: FHWA for pavement restoration, repair, or construction of 
highways that could benefit freight and nonfreight users. $27.5 
billion or 57 percent of DOT's total Recovery Act funding of $48.1 
billion. 

Railroad: Federal Railroad Administration for the High-Speed Intercity 
Passenger Rail program. Such assistance could also enhance freight 
rail where freight and passenger trains share tracks.[A] 
* $8.0 billion or 17 percent of DOT's total $48.1 billion. 

Waterways: Corps for waterways and harbor projects that could benefit 
both freight and nonfreight waterways users. The Corps received 
approximately $4.6 billion in Recovery Act funds. 
* Construction: $637 million.[B]; 
Operations and maintenance: $956 million[C]. 

We also identified around $762 million in TIGER grant funding that 
could benefit freight transportation.[D]. 

EPA Clean Diesel Program funding--$55 million identified for 
multimodal purposes: 

Trucking: To acquire idle reduction and aerodynamic technologies that 
help reduce truck diesel emissions, among other things. 
* $88 million or 30 percent of the total $289 million. 

Railroad: To repower locomotives with emissions reducing technologies. 
* $27 million or 9 percent of the total $289 million. 

Waterways: To repower or retrofit tugboats, marine engines, and 
vessels with emission reducing technologies. 
* $25 million or 9 percent of the total $289 million. 

Build America Bond Program--state and local bonds for 61 
infrastructure projects totaling $18.4 billion: 

Trucking: 
* For 34 highway improvement projects (bond amounts totaled $9.9 
billion). 
* For 19 multipurpose projects including highways (bond amounts 
totaled $7.8 billion). 

Railroad: NA. 

Waterways: For 8 port related projects (bond amounts totaled $782 
million). 

Sources: Federal Highway Administration, Army Corps of Engineers, 
Environmental Protection Agency, and Department of the Treasury. 

[A] Although sharing of tracks with passenger operations can result in 
benefits for freight railroad, it may also impose additional costs if 
the freight railroad is not compensated through cost-sharing 
agreements. 

[B] Amount as of January 2010. 

[C] Amount approved as of August 2009. 

[D] TIGER grants are the Transportation Investment Generating Economic 
Recovery grants. We identified 10 highway grants totaling $188 
million, 4 railroad grants totaling $171 million, and 3 waterways 
grants totaling $48 million that could benefit both freight and 
nonfreight users. We also identified eight intermodal grants totaling 
$355 million that benefited more than one transportation mode. In 
October 2010, DOT awarded nearly $600 million in subsequently 
appropriated (FY 2010) funds to over 70 projects in a second round of 
TIGER grants. 

[End of table] 

[End of section] 

Appendix III: Federal Tax Subsidies and Financing Programs: 

Tax expenditures are revenue losses to the federal government 
resulting from tax provisions, such as federal tax (1) exemptions or 
deductions of the interest earned from certain state and local 
government bonds or (2) credits provided by the Department of the 
Treasury for infrastructure projects. 

We obtained estimates of fiscal years 2003 through 2007 tax 
expenditures from the yearly Analytical Perspectives, Budget of the 
United States Government. The Office of Management and Budget 
aggregates tax expenditures for state and local government bonds used 
to finance government operations, facilities, and services, and also 
identifies the general purpose of the bonds. Thus, we could identify 
the specific transportation mode for which these bonds were used, but 
not necessarily whether the bonds were used for domestic freight or 
nonfreight use. We also used the Statistics of Income's Tax-Exempt 
Bonds articles from the Internal Revenue Service to estimate the 
proportion of state and local government bonds used for transportation 
purposes. 

In addition, three federal financing programs administered by DOT 
provided some subsidies to the freight industry through either 
interest rates or terms that were more favorable than might be 
available in the commercial credit markets. These three programs 
include FHWA's Transportation Infrastructure Finance and Innovation 
Act program,[Footnote 81] the Federal Railroad Administration's 
Railroad Rehabilitation and Improvement Financing program,[Footnote 
82] and the Maritime Administration's ship financing program.[Footnote 
83] We obtained cost estimates from these program offices for fiscal 
years 2003 through 2007 for 28 infrastructure projects that we 
identified as being available for use by freight trucks, freight rail, 
and waterways freight. Table 13 summarizes this information. 

Table 13: Summary of Federal Tax Subsidies and Financing Programs, 
Fiscal Years 2003-2007: 

Federal tax expenditures (totals for fiscal years 2003 through 2007): 

Exemption from federal taxes on interest earned from state and local 
government bonds for general transportation purposes totaled $13.5 
billion for 5 fiscal years. We were unable to estimate specific 
amounts for freight and nonfreight usage or by transportation mode. 

Trucking: The exclusion of interest for financing highway projects and 
rail-truck transfer facilities was $65 million; 
Railroad: The 50 percent tax credit to Class II and III railroads for 
maintenance on their railroad tracks was $340 million; 
Waterways: The tax deferral on capital construction of shipping 
companies was $100 million. Exemption on state and local government-
issued private activity bonds for docks and wharves--including freight 
and nonfreight as well as those for international use--was $681 
million. 

Federal financing programs (totals for fiscal years 2003 through 
2007)[A]: 

Trucking: The Transportation Infrastructure Finance and Innovation Act 
credit assistance program provided $938 million in credit assistance 
to five highway projects available for use by freight and nonfreight 
users at a subsidy cost of about $51 million. TIFIA borrowers also 
obtained other monetary benefited that could not be estimated; 
Railroad: The Railroad Rehabilitation and Improvement Financing 
Program loan and loan guarantee program provided $573 million in 
credit assistance for 18 rail projects.[B] While borrowers paid $17 
million in credit risk premiums to offset the risk of default, 
additional monetary benefits may accrue, but could not be estimated; 
Waterways: Federal tax expenditures (totals for fiscal years 2003 
through 2007): The ship financing program provided loan guarantees to 
3 freight waterway shipping companies for ship construction or 
reconditioning at a subsidy cost of $13 million on guarantees totaling 
$302 million. 

Sources: Department of Transportation; President's Budget; and 
Internal Revenue Service, Statistics of Income Bulletin. 

[A] We use the term subsidy here to refer to any form of financial 
assistance provided to a business or economic sector (see footnote 24) 
and not the subsidy costs of the federal financial programs as defined 
by the Office of Management and Budget. OMB Circular A-11 defines 
subsidy cost as the "estimated present value of the cash flows from 
Government (excluding administrative expenses) less the estimated 
present value of the cash flow to the Government resulting from a 
direct loan or loan guarantee, discounted to the time when the loan is 
disbursed." 

[B] Twenty percent of the program's funding is reserved specifically 
for projects benefiting Class II and III railroads; however, from 
fiscal year 2003 through 2007 all borrowers have been Class II and III 
railroads. 

[End of table] 

[End of section] 

Appendix IV: Freight External Cost Estimates from the Literature: 

Cost per ton-mile (in 2010 cents): 

Type: Congestion delay; 
Trucking[A]: 0.24 to 0.58; 
Railroads: 0.03; 
Waterways: Not estimated. 

Type: Accident; 
Trucking[A]: 0.11 to 2.15; 
Railroads: 0.24; 
Waterways: Not estimated. 

Type: Air pollution, health; 
Trucking[A]: 0.11 to 1.67; 
Railroads: 0.01 to 0.38; 
Waterways: 0.09 to 1.87. 

Type: Climate change; 
Trucking[A]: 0.03 to 2.95; 
Railroads: 0.01 to 0.51; 
Waterways: 0.00 to 0.25. 

Type: Noise; 
Trucking[A]: 0.05; 
Railroads: 0.05; 
Waterways: Not estimated. 

Source: Data reproduced primarily from Mark Delucchi and Don McCubbin, 
"External Costs of Transport in the U.S.", Handbook of Transport 
Economics, eds. by A. de Palma, R. Lindsey, E. Quinet, and R. 
Vickerman (Edward Elgar Publishing Ltd., forthcoming). 

[A] Data are largely representative of intercity freight portion, not 
necessarily local freight. 

[End of table] 

[End of section] 

Appendix V: Policy Options for Addressing Tradeoffs between Efficiency 
and Cost Recovery: 

The following cost recovery options have been identified in the 
economics literature:[Footnote 84] 

* General subsidy. Infrastructure users could be charged for the 
marginal public and external costs they impose, and any shortfall in 
the coverage of total costs could be paid out of general government 
funds. This policy would promote efficient use of existing 
infrastructure; however, it would require higher general fund taxes 
(which cause their own economic distortions) than would otherwise be 
necessary and taxpayers who make little use of infrastructure may 
consider this to be unfair. 

* Ramsey pricing. If infrastructure users can be classified into 
different groups depending on the strength of their demand for 
infrastructure use, then those individuals who would not reduce their 
use significantly, even if they were charged an amount that exceeded 
the marginal costs they impose, could be charged a higher price to 
cover fixed costs. Users with weaker demands could be charged prices 
equal to their marginal costs. Under these conditions, infrastructure 
would be utilized up to an efficient level, even though some users are 
charged more than their marginal costs. The principal impediment to 
implementing this approach is the difficulty of estimating the 
strength of various users' demand. In addition, users with high 
demands may consider it unfair to be charged higher fees than other 
users solely on that basis. 

* Two-part tariffs. Infrastructure users could be charged two types of 
fees. One could be a flat-rate fee to cover fixed costs that everyone 
could pay to gain access to the infrastructure. The second fee could 
be a per-use charge designed to cover the marginal costs arising from 
each use. This policy option could lead to less-than-efficient levels 
of infrastructure use because some who would have used the 
infrastructure if only the per-use fee were charged may not use it if 
the additional access fee were charged. This approach might be made 
more attractive to and be perceived as more equitable by different 
types of users if they were given a choice between (1) a high access 
fee with a low per-use charge and (2) a lower access fee with a higher 
per-use charge. 

* Average-cost pricing. Charging users for the average, rather than 
marginal, costs that they impose would raise sufficient revenue to 
cover all costs; however, this policy would reduce efficiency because 
some users who would use the infrastructure if they were charged only 
for their marginal costs may not be willing to use it if they were 
charged the higher amounts needed to cover average costs. 

[End of section] 

Appendix VI: GAO Contacts and Staff Acknowledgments: 

GAO Contacts: 

Phillip R. Herr, (202) 512-2834 or herrp@gao.gov James R. White (202) 
512-9110 or whitej@gao.gov: 

Staff Acknowledgments: 

In addition to the contacts named above, Andrew Von Ah (Assistant 
Director), James A. Wozny (Assistant Director), Max B. Sawicky 
(Assistant Director), Peace Bransberger, Bertha Dong, Brian James, 
Bert Japikse, Delwen A. Jones, Steve Martinez, Ed Nannenhorn, and 
Donna Miller made key contributions to this report. 

[End of section] 

Footnotes: 

[1] Pipelines are also a freight mode for transport of oil and gas, 
and can compete with other modes, but are not considered in the scope 
of this report. 

[2] Department of Transportation, Bureau of Transportation Statistics, 
U.S. Freight on the Move, Highlights from the 2007 Commodity Flow 
Survey Preliminary Data (Washington, D.C.). 

[3] Ton-miles are determined by multiplying the aggregate weight of 
freight by the distance that weight is carried. 

[4] FHWA, Freight Transportation: Improvements and the Economy, (June 
2004). 

[5] A recent study estimates that about 12 percent of truck ton-miles 
can potentially shift to rail or waterways; see James J. Winebrake and 
James J. Corbett, "Improving the Energy Efficiency and Environmental 
Performance of Goods Movement," in Climate and Transportation 
Solutions: Findings from the 2009 Asilomar Conference on 
Transportation and Energy Policy, edited by Daniel Sperling and James 
S. Cannon (Institute of Transportation Studies, University of 
California, Davis, 2010). 

[6] The Safe, Accountable, Flexible, Efficient Transportation Equity 
Act: A Legacy for Users, Pub. L. No. 109-59, § 1101, 119 Stat. 1144, 
1153 (August 10, 2005). 

[7] Pub. L. No. 109-59, §§ 1305(d)(1) and 1306(d)(3), 119 Stat. 1214- 
1217. 

[8] Tax rates are as follows: fuel tax rate of 24.4 cents per gallon 
for diesel and 18.4 cents per gallon for gasoline; sales tax rate of 
12 percent of retailer's sales price for tractors and trucks over 
33,000 pounds gross vehicle weight and trailers over 26,000 pounds 
gross vehicle weight; heavy vehicle use tax rate varies for vehicles 
55,000 pounds or more, and a tire tax rate, depending on tire size and 
weight, of 9.45 cents (4.725 cents for biasply or super single tires) 
for each 10 pounds of the maximum rated load capacity over 3,500 
pounds. 

[9] GAO, High-Risk Series: An Update, [hyperlink, 
http://www.gao.gov/products/GAO-09-271] (Washington, D.C.: January 
2009). 

[10] For fiscal year 2008, about $8 billion was transferred from the 
general fund to the highway account (Pub. L. No. 110-318, § 1(a)(4), 
122 Stat. 3532 (September 15, 2008); in fiscal year 2009 the transfer 
was $7 billion (Pub .L. No. 111-46, §§1, 123 Stat 1970 (August 7, 
2009); and $14.7 billion was transferred to extend highway programs to 
December 31, 2010 (Pub. L. No. 111-147, § 442, 124 Stat. 71, 94 (March 
18, 2010). See GAO, Highway Trust Fund: Improved Solvency Mechanisms 
and Communication Needed to Help Avoid Shortfalls in the Highway 
Account, [hyperlink, http://www.gao.gov/products/GAO-09-316] 
(Washington, D.C.: Feb. 6, 2009). 

[11] Pub. L. No. 111-5, Title XII, 123 Stat, 115, 206 (February 17, 
2009). See appendix II for more details on the Recovery Act. 

[12] TIGER grant funds are available for all freight modes. See 
appendix II for additional information. 

[13] Consolidated Appropriations Act, 2010, Pub. L. No. 111-117, Div. 
A, title I, 123 Stat 3034, 3035-3037 (December 16, 2009). 

[14] See American Association of State Highway and Transportation 
Officials, Freight Rail Bottom Line Report (Washington, D.C., 2003). 

[15] Association of American Railroads, Great Expectations: Railroads 
and the U.S. Economic Recovery (Washington, D.C., February 2010). 

[16] The Saint Lawrence Seaway Development Corporation, a government- 
owned corporation, develops, operates, and maintains that part of the 
St. Lawrence Seaway within the territorial limits of the United States. 

[17] The Water Resources Development Act of 1986, Pub. L. No. 99-662, 
§ 102, 100 Stat. 4082, 4088 (November 17, 1986) codified at 33 U.S.C. 
§ 2212, established the cost-sharing ratios. Fifty percent of the 
costs associated with construction and major rehabilitation of inland 
waterways is paid with funds appropriated from the Inland Waterways 
Trust Fund; 50 percent is paid from funds appropriated from the 
general fund. Operation and maintenance costs of the inland waterway 
system are paid from funds appropriated from the general fund. 

[18] See Inland Waterways Users Board, 23rd Annual Report to the 
Secretary of the Army and United States Congress (August 2009); and 
National Academy of Public Administration, Prioritizing America's 
Water Resources Investments: Budget Reform for Civil Works 
Construction Projects at the U.S. Army Corps of Engineers (Washington, 
D.C., February 2007). 

[19] 23 C.F.R. Part 658. 

[20] Rail Safety Improvement Act of 2008, Pub. L. 110-432, § 104, 122 
Stat. 4848, 4856 (October 16, 2008) codified as positive law at 49 
U.S.C. § 20157. 

[21] 49 C.F.R. Parts 40, 382 (motor carriers) and 219 (rail carriers), 
and 33 C.F.R. Part 95 (watermen). 

[22] See, e.g., 49 C.F.R. § 1572, dealing with transportation 
credentialing and security threat assessments for maritime and land 
transportation workers. 

[23] 40 CFR parts 86, 92 and 94. 

[24] Some waterway infrastructure, such as terminal infrastructure, is 
often privately owned. 

[25] External benefits can also occur in a market. External benefits 
occur when someone not involved in a particular transaction receives 
some benefit which they did not have to pay compensation to receive. 

[26] Some government tax policies, such as the proposed federal tax 
credit for railroad track maintenance, can work in the opposite manner 
to shift the burden of certain private costs from service providers to 
general taxpayers. 

[27] In the scenarios, the competitive rate of return is not shown 
separately, but is included in the cost figure. 

[28] A subsidy is a form of financial assistance provided to a 
business or economic sector. There are reasons for providing a 
subsidy, such as to provide assistance to low-income individuals or to 
correct market failures. For example, in the context of freight 
transportation, there are situations where subsidizing more fuel-
efficient shippers may result in increased economic efficiency because 
it might shift demand from a high external cost alternative to a lower-
cost alternative. However, not all subsidies arise from intentional 
government policies, and from an economic perspective, subsidies that 
result in lower economic efficiency relative to other alternatives 
would be considered less desirable. 

[29] Freight railroads are allowed to charge differential rates--up to 
a certain point--to shippers depending on whether those shippers have 
other modal alternatives, and thus can recover more of their costs 
from what are known as "captive shippers" to allow them to offer lower 
prices and pass on less of their costs where there is more 
competition. For more on captive shippers and rail rates see GAO, 
Freight Railroads: Updated Information on Rates and Other Industry 
Trends, [hyperlink, http://www.gao.gov/products/GAO-08-218T] 
(Washington, D.C.: Oct. 23, 2007). 

[30] One option, known as "Ramsey pricing" would be to charge a higher 
price to users who are less likely to respond by reducing their use of 
the infrastructure than would other users. This approach would be 
similar to the practice just described for railroads. 

[31] The estimates are based on aggregated data in order to compare 
the modes on a nationwide basis. This level of aggregation obscures 
considerable variations in costs and tax payments across individual 
shipments within each mode. For example, for any rail or waterway 
shipment, truck movement may also be involved at either origination or 
destination points. Consequently, the results should be viewed as 
representing averages across all of the marginal shipments that were 
made under a wide variety of different conditions in a wide variety of 
locations. 

[32] Difficulties in estimating the health and mortality costs 
associated with a ton of specific pollution emissions are a key source 
of uncertainty surrounding our estimates. Nevertheless, available 
studies identify values that we can reasonably characterize as 
conservative. Moreover, we can draw conclusions about the relative 
magnitudes of pollution costs. 

[33] Each mile of driving consumes fuel and reduces a tire's tread 
life and, thereby, increases the taxes a driver pays. Toll payments 
typically are not as directly related to each mile driven; however, a 
driver's total toll payments in a given year are likely to be 
correlated with the number of miles driven. 

[34] Our estimates of infrastructure costs, subsidies, and tax 
payments for the trucking mode are based on data for all single-unit 
and combination trucks (excluding pickup trucks). This population of 
trucks will include a variety of trucks that are not used for freight 
services. Data from the 2002 Vehicle Inventory and Use Survey suggests 
that between 64 percent and 83 percent of total vehicle miles traveled 
(VMT) for single-unit trucks and between 94 percent and 100 percent of 
total VMT for combination trucks was attributable to freight providers 
(depending on whether concrete mixers, dump, and trash and recycling 
trucks are included as freight trucks). In total, between 83 percent 
and 94 percent of VMT for all trucks was attributable to freight 
providers. 

[35] To adjust for the large differences in tonnage moved by different 
modes, we use 1 million ton-miles as our unit of measure for marginal 
increases in services and also as a basis for comparing fixed costs 
across modes. This marginal unit represents an increase in service 
provided by entirely new trips as opposed to increases attained by 
adding tonnage to each trip. In the rail and waterway modes, this 
margin could represent one additional trip, whereas in trucking, it 
would represent many additional trips. Given that our data on marginal 
costs and revenues are averaged across all types of freight truck 
trips, the ratio of marginal costs to marginal revenue for trucks 
would be the same, whether our marginal unit is 1 ton-mile or a 
million ton-miles. However, a truck ton-mile does not necessarily 
represent the same unit of service as a rail or waterway ton-mile. 
Rail and waterway networks allow for less flexibility in shipping 
routes than does the highway system; therefore, it may take more ton-
miles to ship a ton of freight between points A and B by rail or 
waterway than it does by truck. To the extent that rail and waterway 
shipments travel more miles between two given points than do truck 
shipments, the trucking costs in table 3 are overstated relative to 
the other modes on a constant-unit-of-service basis. We could not find 
evidence to suggest that these differences are large enough to change 
the direction or the order of magnitude of our findings. Although VMT 
has been used as the marginal unit in some freight trucking analyses, 
this unit is not useful for cross-mode comparison given the huge 
differences in vehicle sizes across modes. 

[36] Because of uncertainties surrounding truck freight ton-mile 
estimates, we present alternative truck ton-mile estimates in appendix 
I. The sensitivity analysis shows that the comparison across modes 
would not change fundamentally with those alternative values; a change 
in the ton-mile estimate changes both the costs and revenues per ton- 
mile in the same direction. 

[37] We did not consider bridge costs. According to TRB, bridge cost 
generally is regarded to be small relative to other highway costs of 
increased traffic. In contrast to a pavement, which is designed to 
fail eventually as a result of fatigue after a specified number of 
loads pass, a bridge is designed for an extended life span, provided 
that it is not exposed to a single load greater than its load-bearing 
capacity. Our pavement cost estimates were based on government 
spending, rather than on actual pavement damage. To the extent that 
governments did not attempt to repair all pavement damage, these 
estimates understate the actual costs. 

[38] FHWA provided us with updated data (representing costs averaged 
over the years 2000 through 2006) for costs funded by the federal 
government. We updated the truck share of state and local government 
costs by applying percentage share estimates from the 1997 FHWA Cost 
Allocation Study to more recent cost data from Highway Statistics (see 
appendix I for methodology details). The average cost per ton mile 
should be close to the cost per marginal ton-mile because each ton-
mile driven by a particular type of vehicle over a specific stretch of 
road under similar conditions imposes approximately the same pavement 
cost. All of the cost estimates--both average and marginal--cited here 
are averaged across all ton-miles driven by trucks nationwide in a 
given year. 

[39] We did not have separate ton-mile data for single-unit or 
combination trucks. However, we estimated from VMT data by different 
weight classes of trucks that combination trucks carried, on average, 
about three times the load of single units resulting in marginal 
revenues exceeding marginal public infrastructure costs by $12,000 per 
million ton-miles for single unit $3,000 for combination trucks. The 
revenues cover the pavement preservation costs when they are allocated 
to these costs rather than being used to address the marginal external 
costs. 

[40] Although only 20 percent of the program's funding is reserved for 
projects benefiting freight railroads other than Class I railroads, 
for fiscal years 2003 through 2007, borrowers have all been Class II 
and III railroads. 

[41] See Congressional Budget Office, Taxing Capital Income: Effective 
Rates and Approaches to Reform (October 2005), table 2. 

[42] An effective tax rate measures the share that taxes take out of 
the return earned on a specific investment. An effective tax rate and 
a statutory tax rate differ in several ways. The effective rate 
applies to the economic income earned over the life of an investment 
and takes account of several factors beyond just the statutory tax 
rate, including differences between tax depreciation and economic 
depreciation, inflation, and differences in the tax treatment of 
income from debt-financed versus equity-financed investments. 

[43] In 2006 CBO estimated that about 37 percent of assets in the 
trucking industry are owned by businesses that are not subject to the 
corporate income tax. CBO did not make a separate effective tax rate 
estimate for these types of businesses in this category. CBO did 
indicate that the average statutory rate for unincorporated businesses 
was 27 percent, compared to the 35 percent for corporations. We assume 
that the effective tax rates for the average unincorporated trucking 
firm is no greater than 18.2 percent, based on CBO's analysis and our 
belief that smaller unincorporated businesses are likely to use the 
most generous capital allowances available in the tax code (expensing 
of capital expenditures under section 179) for a larger proportion of 
their investments than are corporations. 

[44] Appendix IV contains external cost estimates from the literature. 
Although noise costs are relevant to freight transportation, the 
available information indicates that under most conditions noise costs 
are a fraction of total estimated freight costs, and roughly equal for 
truck, rail, and waterborne freight. The evidence also indicates 
external costs that are not well documented, such as water pollution 
and hazardous material releases, would not alter the observed overall 
imbalance between the modes. 

[45] Available external cost estimates from other sources shown in 
appendix IV indicate that air pollution and climate change from all 
surface freight transportation could be as high as 7.6 cents per ton- 
mile of freight. 

[46] The EPA's benefit estimates are from EPA, Final Rulemaking to 
Establish Light-Duty Vehicle Greenhouse Gas Emission Standards and 
Corporate Average Fuel Economy Standards: Regulatory Impact Analysis, 
EPA-420-R-10-009 (April 2010). See appendix I for more detail on GAO's 
computations. 

[47] According to EPA, fine particulate matter can lodge in the lungs, 
aggravate respiratory conditions such as asthma and bronchitis, cause 
lung damage and premature death, and may even be a cause of cancer. 
Nitrogen oxide is a precursor of ground-level ozone, which can 
contribute to health problems similar to those caused by fine 
particulate matter, although less acute. In addition to physical 
health risks, these pollutants also contribute to haze and reduced 
visibility, and a variety of other environmental impacts. 

[48] Ocean-going vessels involved in coastwise freight movements have 
significantly different performance with respect to emissions than do 
inland waterways vessels. However, data are not available to isolate 
the differences, and thus we do not provide separate estimates in this 
report. 

[49] For more information about estimating freight air pollution, see 
Transportation Research Board, Representing Freight in Air Quality and 
Greenhouse Gas Models, NCFRP Report No. 4 (Washington, D.C., 2010); 
and GAO, Climate Change: The Quality, Comparability, and Review of 
Emissions Inventories Vary Between Developed and Developing Nations, 
[hyperlink, http://www.gao.gov/products/GAO-10-818] (Washington, D.C.: 
July 30, 2010). 

[50] Greenhouse gases trap the sun's heat within the earth's 
atmosphere and contribute to climate change. The dominant greenhouse 
gas emission for the transport sector is CO2, but other important 
manmade greenhouse gases include methane, nitrous oxide, and 
fluorinated gases. 

[51] 40 C.F.R. parts 80, 86, 92, 94, 1033, and 1042. 

[52] Our accident data for freight trucking covers trucks of over 
10,000 pounds gross vehicle weight, and may include dump trucks, 
cement mixers, and garbage/refuse haulers. We selected freight vessels 
that were defined as tows, tugs, ships, or barges as representing 
domestic waterborne freight. 

[53] Economists and other researchers have worked to establish 
specific values for the loss of life and serious injuries. Currently 
DOT uses $6 million in its analysis when determining the Value of a 
Statistical Life, which is defined as the value of improvements in 
safety that result in a reduction by one in the expected number of 
fatalities that a regulatory action provides. 

[54] See Jason D. Lemp and Kara M. Kockelman (2008). "Quantifying the 
External Costs of Vehicle Use: Evidence from America's Top-selling 
Light-duty Models," Transportation Research Part D,; Transportation 
Research Board, Paying Our Way, Estimating Marginal Social Costs of 
Freight Transportation, (1996); Forkenbrock, David J. (1999). 
"External Costs of Intercity Truck Freight Transportation," Transport 
Research part A 33, 505-526; and David J. Forkenbrock, "Comparison of 
external costs of rail and truck freight transportation," Transport 
Research, Part A, 35 (2001): 321-337. 

[55] See, for example, Congressional Research Service, Surface 
Transportation Congestion: Policies and Issues, RL33995 (Feb. 6, 
2008); and Federal Highway Administration, Estimated Cost of Freight 
Involved in Highway Bottlenecks (Nov. 12, 2008). Congestion can also 
add to air pollution and other secondary costs, but we did not find 
separate estimates for these types of effects. 

[56] Beyond estimating the external costs of road freight, this study 
also estimates the congestion costs imposed on highway users by 
freight rail at road crossings. See Michael F. Gorman, "Evaluating the 
public investment mix in US freight transportation infrastructure," 
Transportation Research, Part A 42 (2008): 1-14. 

[57] Truck and waterborne freight carriers may add to congestion that 
affects other carriers within the same mode. However, we do not 
consider congestion costs borne by other carriers within the mode as 
external costs for that mode. Nevertheless, there still may be 
misallocation of freight services or resources, even if these costs 
are not considered an external cost. 

[58] We estimated the state and local revenues attributed to freight 
trucks using yearly share ratios compared to a 2000 ratio. The revenue 
estimate is an average across VMT from 2000 to 2006. See appendix I 
for details. 

[59] Fixed private costs are mostly likely passed on fully to 
consumers under competitive conditions. 

[60] If, on average, combination trucks carried about three times the 
load of single units, then the unpriced fixed costs would be about 
$20,000 per million ton-miles for single-unit trucks and $4,000 for 
combination trucks. 

[61] The last federal highway cost allocation study was issued in 
1997, with an update in 2000. FHWA has a new cost allocation study 
ongoing. 

[62] See GAO, Climate Change: Expert Opinion on the Economics of 
Policy Options to Address Climate Change, [hyperlink, 
http://www.gao.gov/products/GAO-08-605] (Washington, D.C.: May 9, 
2008) for a broader and more detailed discussion of this issue. 

[63] Data from the 2002 Economic Census--Vehicle Inventory and Use 
Survey: Geographic Area Series--suggests that, in total, between 83 
percent and 94 percent of VMT for all trucks was attributable to 
freight providers (depending on whether concrete mixers, dump, and 
trash and recycling trucks are included as freight trucks). Between 64 
and 83 percent of total VMT for single-unit trucks and between 94 and 
100 percent of total VMT for combination trucks was attributable to 
freight providers. 

[64] Highway system preservation costs include pavement resurfacing, 
rehabilitation, and reconstruction. 

[65] Although bridges are designed to withstand the load of the bridge 
itself, the load of the heaviest vehicle using the bridge, plus a 
safety factor--and thus have a negligible marginal cost--usage by 
heavy trucks exceeding the bridge weight limit can cause considerable 
wear and tear. 

[66] According to the FHWA official, historically, the annual rate at 
which obligations were expended were as follows: 27 percent in year 1, 
41 percent in year 2, 16 percent in year 3, 5 percent in year 4, 3 
percent in year 5, 3 percent in year 6, 2 percent in year 7, 2 percent 
in year 8, and 1 percent in year 9. 

[67] Federal highway spending is generally used for capital 
investments. Some federal highway spending is identified for 
maintenance of roads on federal lands, such as National Park Service 
and Bureau of Land Management roads. Spending includes FHWA activities 
funded through the general fund and all other federal agencies, such 
as the Forest Service, Bureau of Indian Affairs, Bureau of Land 
Management, as well as the Highway Trust Fund Mass Transit account. We 
do not estimate federal maintenance spending attributable to freight 
trucking, as the FHWA Highway Cost Allocation Study did not include 
funds directly appropriated from general funds. However the amount of 
freight trucking on special federal roads (i.e., roads in national 
parks) is likely to be minimal. 

[68] See appendix A of FHWA's Guidelines For Conducting A State 
Highway Cost Allocation Study Using the State HCAS Tool for examples 
of routine maintenance costs, including roadway surfacing patching and 
scraping, maintenance of signs, painting, and winter plowing, which, 
for the most part, are the responsibility of state and local 
governments. 

[69] We used the study's projections for 2000 as the basis for the 
shares for state revenues from registrations, licenses, and weight- 
distance taxes. FHWA advised against using the 2000 projections for 
motor fuels taxes, so we used its 1994 data as our starting point for 
those taxes instead. 

[70] Freight rails were subject to a federal fuel tax between fiscal 
years 2000 through 2006 for deficit reduction purposes. Freight rails 
paid 4.3 cents per gallon for diesel in 2000 and 2004; 3.3 cents from 
January 1, 2005 through June 30, 2005; 2.3 cents from July 1, 2005 
through December 31, 2006; and nothing after December 31, 2006. 
Railroad companies are subject to the 0.1 cent per gallon Leaking 
Underground Storage Tank tax on motor fuels. 26 U.S.C. § 
4041(a)(C)(ii)(III). 

[71] Transportation Research Board, Paying Our Way, Estimating 
Marginal Social Costs of Freight Transportation (1996) shows 0.03 
cents per ton-mile plus 3 cents per ton per lock passage, converted to 
2010 dollars by GAO. Congressional Budget Office, Paying for Highways, 
Airways, and Waterways: How Can Users Be Charged? (May 1992) shows 
0.06 cents per ton-mile, converted to 2010 dollars by GAO. 

[72] See Congressional Research Service, Harbor Maintenance Trust Fund 
Expenditures (January 25, 2010). 

[73] U.S. Army Corps of Engineers, Navigation Cost Allocation Study 
(1980). The purpose of this study was to determine the feasibility of 
specifically identifying expenditures made exclusively for commercial 
navigation and, where multipurpose features are involved, allocating 
costs among various beneficiaries. 

[74] The Freight Analysis Framework (FAF) integrates data from a 
variety of sources to create a comprehensive picture of freight 
movement among states and major metropolitan areas by all modes of 
transportation. With data from the 2007 Commodity Flow Survey and 
additional sources, FAF version 3 provides estimates for tonnage and 
value, by commodity type, mode, origin, and destination for 2007, the 
most recent year, and forecasts through 2040. 

[75] DOT officials could not provide a definitive explanation as to 
why the BTS data do not capture the full complement of domestic truck 
freight activity. One official said he thought that part of the 
explanation is because the BTS figures do not fully capture the 
movement of some imported commodities. Another DOT official said that 
the BTS figures undercount retail-to-retail or other intercompany 
movements for large retailers that manage their own fleet of truck 
freight vehicles. This official also thinks that BTS undercounts 
freight movements in the 50-to 200-mile range when moving from origin 
to destination. 

[76] In fact, as previously mentioned, some cost and revenue data were 
derived from populations that slightly exceeded all freight trucks. 

[77] DOT plans to release revised estimates for years 1997 and 2002 
based on the new methodology in 2011. 

[78] According to an EPA official, the estimated damages (or 
"disbenefits" in EPA's usage) should not be considered completely 
synonymous with costs because part of the estimate was determined by 
surveying the population on what they would be willing to pay in order 
to extend their health and life by reducing pollution. It is plausible 
that if these respondents were asked how much these pollutants were 
costing them in health and quality of life costs, then they would 
likely have a different response. 

[79] Pub. L. No. 111-5, 123 Stat. 115 (February 17, 2009). 

[80] Recovery Act, Div. B, title I, § 1531. 

[81] Transportation Infrastructure Finance and Innovation Act of 1998, 
Pub.L. 105-178, Title I, Subtitle E, Ch. 1, §§ 1501 to 1504, 112 Stat. 
241 (June 9, 1998; codified as positive law at 23 U.S.C. chapter 6). 

[82] Pub.L. 94-210, Title V, § 502, as added Pub.L. 105-178, Title 
VII, § 7203(a)(1), 112 Stat. 473 (June 9,1998; codified at 45 U.S.C. 
chapter 17). 

[83] 46 U.S.C Chapter 537. 

[84] See, for example, Congressional Budget Office, Paying for 
Highways, Airways, and Waterways: How Can Users Be Charged? (May 1992). 

[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 Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, 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 Web site, 
[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. 

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

Contact: 

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

Congressional Relations: 

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 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, D.C. 20548: