Surface Transportation Congestion: Policy and Issues

Surface Transportation Congestion:
Policy and Issues
Updated February 6, 2008
William J. Mallett
Specialist in Transportation Policy
Resources, Science, and Industry Division

Surface Transportation Congestion: Policy and Issues
Surface transportation congestion most likely will be a major issue for Congress
as it considers reauthorization of the Safe, Accountable, Flexible, Efficient
Transportation Equity Act — A Legacy for Users (SAFETEA), P.L. 109-59, which
is set to expire on September 30, 2009. By many accounts, congestion on the
nation’s road and railroad networks, at seaports and airports, and on some major
transit systems is a significant problem for many transportation users, especially
commuters, freight shippers, and carriers. Indeed, some observers believe congestion
has already reached crisis proportions. Others are less worried, believing congestion
to be a minor impediment to mobility, the by-product of prosperity and accessibility
in economically vibrant places, or the unfortunate consequence of over reliance on
cars and trucks that causes more important problems such as air pollution and urban
sprawl. Trends underlying the demand for freight and passenger travel — population
and economic growth, the urban and regional distribution of homes and businesses,
and international trade — suggest that pressures on the transportation system are
likely to grow substantially over the next 30 years.
Although transportation congestion continues to grow and intensify, the problem
is still geographically concentrated in major metropolitan areas, at international trade
gateways, and on some intercity trade routes. Because of this geographical
concentration, most places and people in America are not directly affected by
transportation congestion. Consequently, in recent federal law, Congress, for the
most part, has allowed states and localities to decide the relative importance of
congestion mitigation vis-a-vis other transportation priorities. This has been
accompanied by a sizeable boost in funding for public transit and a more moderate
boost in funding for traffic reduction measures as part of a patchwork of relatively
modest federally directed congestion programs.
Congress may decide to continue with funding flexibility in its reauthorization
of the surface transportation programs. States and localities that suffer major
transportation congestion would be free to devote federal and local resources to
congestion mitigation if they wish. Similarly, congestion-free locales would be able
to focus on other transportation-related problems, such as connectivity, system
access, safety, and economic development. Alternatively, Congress may want to
more clearly establish congestion abatement as a national policy objective, given its
economic development impact, and take a less flexible and, in other ways, more
aggressive approach to congestion mitigation. Three basic elements that Congress
may consider are (1) the overall level of transportation spending, (2) the prioritization
of transportation spending, and (3) congestion pricing and other alternative ways to
ration transportation resources with limited government spending.
Congress also may want to consider the advantages and disadvantages of
specific transportation congestion remedies. Hence, this report discusses the three
basic types of congestion remedies proposed by engineers and planners: adding new
capacity, operating the existing capacity more efficiently, and managing demand.

Issues for Congress................................................3
Transportation Spending Levels..................................6
Transportation Spending Priorities................................8
Congestion Pricing and Other Alternative Ways to Ration Resources....12
Brief History of Transportation Congestion............................15
Highway Transportation........................................15
Public Transportation..........................................17
Freight Transportation.........................................17
Legislative History of Transportation Congestion........................20
Intermodal Surface Transportation Efficiency Act of 1991
(P.L. 102-240)...........................................20
National Highway System Designation Act of 1995 (P.L. 104-59).......23st
Transportation Equity Act for the 21 Century (P.L. 105-178;
P.L. 105-206)...........................................23
Safe, Accountable, Flexible, Efficient Transportation Equity Act —
A Legacy for Users (P.L. 109-59)............................24
Transportation Congestion: Concepts, Measures, and Trends...............27
Measures and Trends in Road Traffic Congestion....................29
Current Trends in Road Traffic Congestion....................32
Interurban Road Traffic Congestion..........................36
Road Bottlenecks.........................................36
Road Congestion at International Gateways....................37
Measures and Trends of Congestion in Public Transit................38
Measures and Trends of Congestion in Rail........................41
Freight Rail Congestion Measures............................41
Trends in Freight Rail Congestion............................42
Intercity Passenger Rail (Amtrak) Congestion Measures..........45
Intercity Passenger Rail (Amtrak) Congestion Trends............45
The Costs of Transportation Congestion...............................46
Transportation Congestion Remedies.................................48
Operating Existing Capacity More Effectively......................50
Managing Demand............................................50
Congestion Pricing........................................51
Land Use Strategies.......................................52
Institutional Issues............................................52
Expanding Rail Capacity.......................................53
Intermodalism in Freight Transportation...........................54
Concluding Observations...........................................54

Figure 1. Motor Vehicle Travel and Road Capacity, 1941-2005............16
Figure 2. U.S. Cost of Logistics, 1984-2005 (percentage of Gross Domestic
Product) ....................................................20
Figure 3. The Relationship Between Speed and Vehicle Flow on Freeways...32
Figure 4. Proximate Causes of Road Traffic Congestion..................33
Figure 5. Road Traffic Congestion, 1982-2005.........................34
Figure 6. U.S. Merchandise Trade by Region, 1980-2005.................38
Figure 7. Transit Vehicle Utilization, 1995-2004........................41
Figure 8. Freight Rail Traffic Density, 1980-2006.......................43
Figure 9. Average Speed of Freight by Rail, 1980-2006..................44
Figure 10. Average Freight Rates, 1980-2006
(constant 2000 cents)..........................................45
List of Tables
Table 1. SAFETEA Authorization Levels, by Legislative Titles and Selected
Programs, FY2005-FY2009....................................26
Table 2. Top 10 Metropolitan Areas by Transit Usage, 2004...............40

Surface Transportation Congestion:
Policy and Issues
Transportation congestion most likely will be a major issue for Congress as it
considers reauthorization of the Safe, Accountable, Flexible, Efficient Transportation
Equity Act — A Legacy for Users ( SAFETEA), P.L. 109-59, which is set to expire
on September 30, 2009. By many accounts, congestion on the nation’s road and
railroad networks, at seaports and airports, and on some major transit systems is a
significant problem for many transportation users, especially commuters, freight
shippers, and carriers. Moreover, trends underlying the demand for freight and
passenger travel — population and economic growth, the urban and regional
distribution of homes and businesses, and international trade — suggest that
pressures on the transportation system are likely to grow in the years ahead.
A number of experts and organizations believe that congestion has reached
crisis proportions. In announcing a new National Congestion Strategy in May 2006,
then Secretary of Transportation Norman Mineta stated that “congestion is one of the
single largest threats to our economic prosperity and way of life.”1 In a similar vein,
the Transportation Research Board (TRB) currently has congestion on its “critical
issues” list as one of the most pressing problems of the transportation system, arguing
“if the 20th century can be called the era of building, the 21st may be called the era of
congestion.”2 More recently, in January 2007, the U.S. Government Accountability
Office (GAO), for the first time, placed transportation financing and capacity on its
list of high-risk federal programs and operations.3
Not everyone agrees that congestion is a major, national problem. Some see it
as a minor impediment to mobility, others as an unfortunate by-product of prosperity
and accessibility in economically vibrant places. Several environmental groups argue
that congestion is less the problem than the over reliance on the cars and trucks that
cause it. Indeed, this over reliance on highway transportation, they believe, leads to
more important problems, such as suburban sprawl and air pollution. Furthermore,
because the problem is geographically concentrated, most places and people in
America do not suffer noticeable levels of congestion. Thus, many might question
to what extent transportation congestion is a national problem warranting a federal
government response. In uncongested regions, transportation problems are more

1 U.S. Department of Transportation, “National Strategy to Reduce Congestion on
America’s Transportation Network,” May 2006, at [
conginitove rview070201.htm] .
2 Transportation Research Board, Critical Issues in Transportation (Washington, DC,

2006), p. 2, at [].

3 U.S. Government Accountability Office, High Risk Series: An Update, GAO-07-310,
January 2007, at [].

often to do with basic connectivity of the transportation system, system access, and
economic development.
Connectivity, system access, economic development, and congestion relief are
some of the objectives of national transportation policy that also include mitigating
the negative effects of transportation, such as deaths, injuries, and environmental
damage. According to 49 U.S.C. § 101,
The national objectives of general welfare, economic growth and stability, and
security of the United States require the development of transportation policies
and programs that contribute to providing fast, safe, efficient, and convenient
transportation at the lowest cost consistent with those and other national
objectives, including the efficient use and conservation of the resources of the
United States.
To accomplish these objectives, the federal government regulates transportation
activities and provides funding to encourage states and local governments to build
and operate transportation infrastructure. Since the beginnings of this “federal-aid”
system, there have been major debates about how these funds should be distributed
and spent. An underlying tension throughout these debates has been whether to
distribute funds to encourage the pursuit of nationally defined transportation goals,
such as the building of the Interstate system, or to distribute funds equally between
the states (according to a predefined formula) and allow them to pursue their own
objectives.4 In SAFETEA, about 90% of highway funds are authorized to be
distributed by formula, and states are guaranteed by FY2008- FY2009 a 92% return
on money paid into the highway account of the Highway Trust Fund.5
Because transportation congestion is geographically concentrated, Congress has
tended to favor a state and local approach to solving transportation congestion in the
recent history of the federal surface transportation program. This has been
accompanied by several sizeable boosts in funding for public transit and traffic
reduction measures directed to major metropolitan areas in an attempt to curb the
negative effects of cars and trucks, including road traffic congestion. Congress also
has enacted a patchwork of other programs to deal with congestion at the national
level, with some success, but these have generally been relatively modest efforts.
Consequently, the flexibility provisions of recent federal laws, and with them the
equity provisions that attempt to return to each state the taxes paid by its highway
users into the highway account of the Highway Trust Fund, have largely left it to the
states, and in some cases metropolitan planning organizations, to decide funding
The extent to which Congress decides congestion is a national problem to be
solved by federal dictates, and funding may be a major issue in reauthorization.
Congress may decide its current “bottom-up” approach to planning and programming
transportation improvements, with some modifications, is the best approach to

4 See U.S. Department of Transportation, Federal Highway Administration, America’s
Highways, 1776-1976 (Washington, DC, 1976), especially Part Two, Chapter 1.
5 Transportation Weekly, “Congress Completes Work on Highway Bill,” vol. 6, issue 34,
August 4, 2005.

congestion in the broader scheme of transportation priorities. Conversely, Congress
may decide that congestion warrants a stronger role for the federal government.
Three broad elements of the issue are discussed here: overall levels of transportation
spending, the prioritization of transportation spending, and congestion pricing and
other alternative rationing schemes that require limited government spending.
Although congestion is being experienced throughout the transportation system,
including at ports and airports, this report is limited to a discussion of congestion
associated with the surface transportation system — highways, public transit, and
freight and passenger rail. Because these modes connect with ports and airports,
there is some discussion of intermodal issues at these nodes as well, but the report
does not discuss congestion in the waterway or airway systems per se. This report
begins by outlining in broad terms some of the issues that Congress may face in the
reauthorization debate. This is followed by a brief history of transportation
congestion in the United States, and how Congress has dealt with the issue in the
recent past. It then goes on to discuss transportation congestion concepts, measures,
and trends, followed by information on the national costs of congestion. The final
section lays out some of the major types of congestion remedies that have been
proposed by transportation engineers, planners, and policy makers.
Issues for Congress
Most experts agree that surface transportation congestion has grown over the
past few decades and, moreover, that the demand for surface transportation services
is likely to continue growing over the next few decades. According to one national
assessment of highway congestion by the Texas Transportation Institute (TTI), total
delay in 437 urban areas increased five-fold between 1982 and 2005, and delay per6
peak-period traveler almost tripled. Anecdotal evidence suggests that overcrowding
is a growing problem in some major transit systems and that conflicts between freight
and passenger rail trains (commuter and intercity) are an issue for both. In the freight
rail industry, the Congressional Budget Office (CBO) notes that average speeds, one
indicator of congestion, are lower now than at anytime since the early 1980s except
for the 1997-1998 period following the merger of Union Pacific and Southern7
Pacific. With dramatic increases in foreign trade, many fear that ports and border
crossings have become significant bottlenecks to the flow of commerce.
Despite these trends, the question remains as to whether or not congestion is a
national problem and, therefore, should be a specific goal of national transportation
policy. Although congestion has intensified and spread, congestion is geographically
concentrated in major metropolitan areas, at international trade gateways, and on
some intercity trade routes. Because of this geographical concentration, most states
and localities do not suffer any appreciable transportation congestion directly.
Moreover, some argue that even in places with relatively intense congestion

6 Texas Transportation Institute, Urban Mobility Report 2007 (College Station, TX, 2007),
at [].
7 U.S. Congressional Budget Office, Freight Rail Transportation: Long-Term Issues,
January 2006, at [].

problems, it only adds a few extra minutes to daily travel and that many actually
enjoy the extra time alone in the car away from the pressures of work and family.8
Seen in terms of an entire trip, including the time it takes to park and walk to the
office, one expert believes the extra time caused by freeway delay is relatively
minor.9 Some even go so far as to suggest that much like a crowded restaurant or
nightclub, congestion is a sign of success and its costs must be balanced against the
benefits of access to jobs, stores, recreational amenities, etc. that congested regions
provide.10 Environmental organizations generally argue that road traffic congestion
results from an unbalanced transportation system, one that favors cars and trucks, and
that urban sprawl, air pollution, and noise, not road traffic congestion per se, should
be the focus of national policy.11
The alternative view is that transportation congestion is a major problem,
national in scope, and, if unchecked, a problem that will intensify and spread over the
next 25 years. Many experts point out that although congestion may be highly
localized, because transportation is a network that serves the U.S. population in a
variety of ways, its economic effects are national. Most obviously, freight movement
is largely dependent on a national transportation network in which a bottleneck in one
place, such as southern California, may affect businesses and consumers in largely
congestion-free Nebraska. Moreover, these experts point out the national network
effects are becoming increasingly important as supply chains lengthen and become
more complex. Similarly, although passenger transportation is mostly a local affair,
congestion on roads that service airports and other passenger terminals may also
result in inefficient intercity passenger travel, dragging down the productivity of
businesses that rely on it for managing far-flung operations.
Local congestion may also be thought of as a national issue in that the places
where it is found tend to be the hubs of the national economy and its costs, therefore,
are not inconsequential in terms of the national economy. For instance, the 28
metropolitan areas that experienced 40 hours or more of annual delay per peak-period
traveler (as measured in 2005 by TTI) account for more than 45% of total personal
income in the United States (in 2005).12 Most businesses rely, to one degree or
another, on the efficient transportation of people locally, whether it is the
transportation of managers to business meetings, workers to work, or customers to
places where products are consumed. Research has shown that metropolitan areas

8 Downs, Anthony, Still Stuck in Traffic: Coping with Peak-Hour Traffic Congestion
(Washington, DC: Brookings Institution Press, 2006).
9 Taylor, Brian D., “Rethinking Congestion,” Access, vol. 21 (2002), pp. 8-16.
10 Ibid.; Downs, 2006; El-Geneidy, Ahmed M. and David M. Levinson, “Access to
Destinations: Development of Accessibility Measures,” report prepared for the Minnesota
Department of Transportation, May 2006, at [].
11 See, for instance, Sierra Club, Highway Health Hazards (San Francisco, CA, 2004), at
[ ht t p: / / er r acl ub.or g/ spr a wl / r epor t 04_hi ghwayheal t h/ r epor t .pdf ] .
12 CRS calculation based on U.S. Bureau of Economic Analysis, “Personal Income for
Metropolitan Areas, 2006,” Table 1, News Release, August 7, 2007, at
[] .

with the largest labor markets tend to have the highest productivity.13 Consequently,
when added together, the local costs of congestion, some argue, are significant in
national terms.
Another commonly expressed view is that given current trends in the supply and
demand for transportation the problems of congestion will affect more people and
more businesses in the future. Road traffic congestion, for instance, is growing
fastest in the smaller urban areas included in the TTI study, though admittedly from
a small base. However, research by the Federal Highway Administration (FHWA)
shows a wider problem when it projects future demand on the current highway
system.14 Underlying these trends are broader trends in population and the economy.
For example, the population is expected to reach 364 million by 2030, an increase of
about 20% from 2007.15 Over the same period, the CBO projects GDP to increase
by about 70% (in real terms).16 Furthermore, the FHWA predicts that freight
movements will nearly double between 2002 and 2035.17
The federal surface transportation program approach to congestion tends to view
it as a state and local issue, not as a major national problem. At least as far back as
passage of the Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991
(P.L. 102-240), Congress has tended to leave to the discretion of the states, within
certain planning parameters, the relative weight to be placed on congestion mitigation
vis-a-vis other transportation priorities. In this regard, many argue that governments
and other stakeholders closest to transportation problems are in the best position to
craft solutions. Another issue since the 1980s, with the near completion of the
Interstate system, has been the controversy regarding state payments to and from the
Highway Trust Fund (HTF), known as the “donor-donee” debate.18 This debate
focuses on the perceived fairness of the relative size of each state’s payments to and
receipts from the highway account of the Highway Trust Fund. Increasingly over the
years, federal law has attempted to equalize these amounts rather than concentrate
funding where needs are greatest. Several new federal programs to tackle congestion

13 Crafts, Nicholas and Timothy Leunig, “The Historical Significance of Transport for
Economic Growth and Productivity,” background paper for the Eddington Transport Study,
October 2005, at [
14 See the maps in U.S. Department of Transportation, Federal Highway Administration,
Freight Facts and Figures 2007 (Washington, DC, 2007), pp. 31-32, at
15 U.S. Census Bureau, Statistical Abstract of the United States, 2008 (Washington, DC,

2007), p. 8, at [].

16 U.S. Congressional Budget Office, The Long Term Budget Outlook: Supplemental
Datasheet (Washington, DC, December 2007), at
[ ht t p: / / f t pdocs/ 88xx/ doc8877/ Suppl ement a l Dat a.xl s] .
17 Federal Highway Administration, 2007, p. 11.
18 See CRS Report RL31735, Federal-Aid Highway Program: “Donor-Donee” State Issues,
by Robert S. Kirk.

nationally have been developed, but, in dollar terms, these have been relatively
Because state and local funding flexibility has been a significant feature of
federal transportation policy since ISTEA, Congress may decide to continue with this
approach in reauthorization. States and localities that suffer major transportation
congestion would be free to devote federal and local resources to congestion
mitigation if they wish. Similarly, congestion-free locales would be able to focus on
other transportation-related problems, such as connectivity, system access, safety, and
economic development. Alternatively, Congress may want to take a less flexible and,
in other ways, more aggressive approach to congestion mitigation. Three basic
elements to the problem that Congress may want to consider are (1) the overall level
of transportation spending, (2) the prioritization of transportation spending, and (3)
congestion pricing and other alternative ways to ration transportation resources.19
Transportation Spending Levels
The amount of federal funding for surface transportation programs is a major
issue during all reauthorization debates and will undoubtedly be an issue in the
reauthorization of SAFETEA. Some observers contend that America is
underinvesting in transportation infrastructure, resulting in deteriorating conditions
and worsening performance, including growing congestion.20 One alternative to
addressing transportation congestion, in this view, is a significant increase in the
overall level of infrastructure investment to deal with the existing backlog of projects
and future needs. The most recent needs assessment by the U.S. Department of
Transportation (USDOT) suggests that the cost to maintain the current condition and
operational performance of the highway system is about 12% more annually than is
being currently spent by all levels of government. For transit, the figure is 25%.
Spending to improve conditions and reduce congestion would be greater than this.21
It should be pointed out that, as with any attempt to estimate current and future
system conditions and performance, there are a host of simplifying assumptions,
omissions, and data problems that influence the results. Nevertheless, this analysis
suggests that if total government spending is not increased above current levels, the
physical condition of system elements may decline and congestion, particularly
highway congestion, will continue to increase.
An alternative view of the overall level of government transportation spending
is that it has not been dramatically deficient. In this view, deteriorating performance,
and in some places deteriorating conditions, are the result of resources not being

19 For a discussion of these three elements in the early 1990s, see CRS Report 93-107,
Transportation Infrastructure: Economic Issues and Public Policy Alternatives, by J.F.
Hornbeck. (Out of print; available from the author.)
20 See, for instance, American Society of Civil Engineers, “Report Card for America’s
Infrastructure 2005,” at [].
21 U.S. Department of Transportation, Federal Highway Administration and Federal Transit
Administration, 2006 Status of the Nation’s Highways, Bridges, and Transit: Conditions and
Performance (Washington, DC, 2007), at [

directed to the parts of the system that are in greatest demand and, therefore, have the
greatest needs for maintenance and expansion. Indeed, USDOT’s own analysis of
historic spending patterns shows that total government spending in highways and
transit, including capital spending, has generally kept pace with usage since the early
1980s, although the federal share has declined. Capital spending by all levels of
government on highways per vehicle mile has remained relatively constant since
about 1980, at around 2.5 cents per vehicle mile (in real terms).22 Over this period,
the federal share declined from close to 60% to a little under 40% at the end of the

1990s, but has since rebounded to about 44% in 2004.23

In terms of the nation’s transit systems, the USDOT analysis shows that total
government spending on capital and operations grew by approximately 80% between
1980 and 2004 (in real terms), much faster than passenger trips, which grew by
12%.24 The federal share of total spending declined from 42% to 25% over this
period.25 The federal share of capital spending in 2004 was 39%, somewhat lower
than the approximately 50% share that existed in the mid-1990s.26 In 2004, the
federal government funded about $36 billion of highway and transit capital
expenditure, with 86% going to highways and 14% to transit.27 The transit share
increases to about 16% if all government spending is included.28
Consequently, assessments of highways nationally reveal that conditions have
generally improved overall during the past decade, particularly in rural areas, but
have declined in large urban areas.29 Similarly, bridge conditions have improved, but
to a much greater extent in rural areas than in urban areas.30 As noted above,
operational performance on the urban highway system has generally declined, but
there are also growing pressures on the higher elements of the rural highway system,
especially rural interstates.31 Transit conditions and performance have remained
about the same over the past decade, but rail system performance has declined to
some extent.32

22 Ibid., exhibit 6-11.
23 Ibid., exhibit 6-8. The federal share of highway spending as a whole is lower, currently
about 22% of spending by all levels of government. The same pattern of a shrinking federal
share since 1980 is similar, however.
24 Ibid., exhibit 6-22; American Public Transportation Association, “Unlinked Passenger
Trips by Mode, 1890-2004,” at [].
25 Ibid., exhibit 6-20.
26 Ibid., exhibit 6-23.
27 Ibid., exhibits 6-8, 6-23.
28 Ibid., exhibits 6-8, 6-20.
29 Ibid., exhibit 3-4.
30 Ibid., exhibit 3-18.
31 Ibid., exhibit 4-12.
32 Ibid., exhibits 3-24, 3-28, 4-15, 4-18.

Some experts, however, believe that investment in the freight rail industry fell
behind demand at some point over the past decade or so, leading to rail congestion
and higher prices for shippers.33 Freight rail, as a predominantly private industry,
depends on investment received mostly from railroad profits or from money
borrowed in capital markets to be paid back with future revenues. One view is that
these sources of investment will be adequate to cope with future demand. Another
view is that because of the great risks inherent in investing in rail infrastructure and
the demands of shareholders, the railroads themselves will not be able to supply the
necessary capital to expand capacity. In that case, some contend that government
financial assistance will be needed, otherwise rail congestion will grow and more
freight will be diverted to the roads.34
As the case of the railroads reminds us, not all transportation infrastructure
investment comes from federal, state, and local government. The private sector is a
major source of investment and not just in rail transportation. A flurry of recent
major privatization efforts, such as the Chicago Skyway and the Indiana East-West
Toll Road, have increased interest in this approach. Thus, some argue that there is
a need for much greater investment in transportation, but that the federal government
should consider using its resources to leverage private investment through public-
private partnerships. Others argue that these types of public-private partnerships will
be limited to only a few places with the highest profit potential and that investment
could be quickly cut off if macroeconomic conditions change.
Transportation Spending Priorities
With growing pressure on transportation infrastructure but competing claims on
governmental resources, another issue for congressional consideration is improving
the efficiency of federal investments. Some argue that prioritizing investments may
be a better way to deal with congestion mitigation than the scattershot, “more-is-
better” approach. Several aspects of prioritizing federal transportation spending to
mitigate transportation congestion could be of interest to Congress. These are
prioritizing projects by location and project type, and the issue of mode-neutrality.
Inherent in these discussions, of course, is how project decisions are made and the
ways in which the relationships between federal, state, and local governments affect
the outcome. This is another aspect of prioritization that may be of interest to
Continued federal transportation funding likely will be needed to maintain and
operate the transportation system as a whole and to meet other national transportation
goals such as rural access, urban mobility, safety, and national security. However,
it can be argued that if mitigating congestion in the name of enhancing national
mobility and economic productivity is a national goal, then federal funding will need

33 Testimony of Carl D. Martland, Senior Research Associate, Massachusetts Institute of
Technology, in U.S. Congress, House Committee on Transportation and Infrastructure,
Subcommittee on Railroads, U.S. Rail Capacity Crunch, April 26, 2006.
34 American Association of State Highway and Transportation Officials (AASHTO),
Transportation, Invest in America: Freight-Rail Bottom Line Report (Washington, DC,

2003), at [].

to be focused in the places that promise the greatest return: those with the most
congestion. The three major locales of transportation congestion are major
metropolitan areas, some intercity trade routes, and foreign trade gateways.
An oft-cited argument for targeting federal resources toward congested places
is that while the project costs of congestion mitigation are local, the benefits, at least
in part, are regional or national in scope. In addition, fixing transportation
bottlenecks is very often a hugely expensive proposition and, therefore, beyond the
means of a single locality or state. Moreover, many point out that in addition to the
pecuniary costs of large transportation facilities, costs associated with local
environmental and social disruptions must be mitigated.
Another aspect of prioritizing federal funding to mitigate congestion is the way
in which projects are planned and funded within states and regions. For the most
part, project development and funding decisions are made by state departments of
transportation (DOTs). Metropolitan planning organizations (MPOs) have assumed
a greater role over the years, but not enough to fundamentally change the traditional
federal-state intergovernmental relationship that has existed since the beginning of
the Federal-Aid Highway Program.35 One effect of this, some have suggested, is that
highway funding tends to be funneled disproportionately toward rural areas at the
expense of urban and suburban areas where needs, including congestion mitigation
needs, are greatest. A study of Ohio found this to be the case because many
municipal roads are ineligible for state funding, state gas taxes are limited by state
law to highway projects, and state apportionments are made equally to counties
without regard to needs such as population, miles of road, and traffic volumes.36
Some of the same processes may also occur within metropolitan regions that
comprise many local jurisdictions. For instance, some observers contend that local
government officials are often more concerned about receiving their “fair share” of
funding than they are about solving regional problems such as transportation
congestion. MPOs also tend in most instances to be dominated by suburban areas at
the expense of center cities because voting power is often not weighted by population
size.37 Of course, weighted voting is no guarantee that a central city will not be
dominated by surrounding jurisdictions when collectively they comprise a larger
share of the regional population.

35 Puentes, Robert and Linda Bailey, “Increasing Funding and Accountability for
Metropolitan Transportation Decisions,” in Bruce Katz and Robert Puentes, eds., Taking the
High Road: A Metropolitan Agenda for Transportation Reform (Washington, DC:
Brookings Institution Press, 2005).
36 Edward Hill et al., “Slanted Pavement: How Ohio’s Highway Spending Shortchanges
Cities and Suburbs,” in Katz and Puentes, 2005.
37 Downs, Anthony and Robert Puentes, “The Need for Regional Anticongestion Policies,”
in Katz and Puentes, 2005; Lewis, Paul G., “Regionalism and Representation: Measuring
and Assessing Representation in Metropolitan Planning Organization,” Urban Affairs
Review, vol. 33, no. 6 (July 1998), pp. 839-853; Association of Metropolitan Planning
Organizations, “AMPO Survey Results: Policy Board Structure,” at
[ ht t p: / / www.ampo.or g/ asset s / 62_pol i c yboar dst r uct ur e.doc] .

A number of other factors have also been found to affect transportation
investment decisions.38 Broad stakeholder involvement requirements in federal law
and, in some cases, the need for local voter approval can have a major influence on
which types of projects move forward and which do not. For example, freight
interests, a relatively minor constituency, argue that such requirements often lead to
the prioritization of passenger projects over freight projects. In addition, state and
local officials, needing to forge consensus on major investment decisions, tend to
favor system preservation, maintenance, and operations projects because they are
comparatively easy and quick to implement. By contrast, major capacity expansion
projects are typically controversial and can take a decade or two to complete. Added
to this is the fact that densely populated urban areas often have limited space
available for major new infrastructure and that old and inadequate infrastructure can
be very difficult and expensive to expand.
Choosing among the types of strategies that provide the most cost-effective
reductions in congestion could be done in a number of ways. The most effective
projects are likely to vary from place to place and situation to situation, requiring
local solutions rather than national dictates. However, Congress may require project
alternatives to be chosen after an assessment of the full benefits and costs, with
congestion mitigation and economic efficiency as high priorities.39 A major study of
transportation in the United Kingdom found that projects aimed at relieving
congestion “offer remarkably high returns, with benefits four times in excess of costs
on many schemes, even once environmental costs have been factored into the
assessment.”40 A different approach is a performance-based assessment in which a
federal standard or goal is set, such as a certain level of congestion reduction, freeing
state and local governments to determine the most efficient way of meeting the
go a l . 41
Another important issue with respect to prioritization is “mode neutrality.”
Traditionally, federal surface transportation funding has been focused on highways
and transit. This has made it difficult to fund projects involving modes that fall
outside these categories, such as freight rail or multi-modal projects.42 Program
changes have been made over the years to allow greater flexibility, but some argue

38 U.S. General Accounting Office, Surface Transportation: Many Factors Affect Investment
Decisions, GAO-04-744 (Washington, DC, June 2004), at [
39 U.S. General Accountability Office, Highway and Transit Investments: Options for
Improving Information on Projects’ Benefits and Costs for Increasing Accountability for
Results, GAO-05-172 (Washington, DC, January 2005), at [
40 HM Treasury and Department for Transport, The Eddington Transport Study, Executive
Summary (London, December 2006), p. 6, at [


41 Cox, Wendell, Alan E. Pisarski, and Ronald D. Utt, “Rush Hour: How States Can Reduce
Congestion Through Performance-Based Transportation Programs,” Heritage Foundation
Backgrounder, no. 1995 (January 10, 2007), at [
42 U.S. General Accounting Office, June 2004.

that these changes have not gone far enough. An opposing view is that when private
transportation infrastructure providers are involved, it is very difficult if not
impossible to properly assess the public benefits and costs of public subsidies.
Others fear that subsidizing private businesses may substitute public investment for
private investment with no net gain for the transportation system, or that such
assistance may provide some businesses an unfair advantage over others.
Mode neutrality in transportation congestion mitigation is still an issue in the
relative balance between funding highways and transit. Some argue that highway
congestion cannot be solved by building more highway capacity or otherwise
improving service because this only encourages or “induces” more people to travel
by highway, thereby restoring the same, or an even higher, level of congestion.
Instead, they contend that alternatives such as public transit in concert with land use
measures to encourage the use of alternative modes of travel are the only way around
congestion.43 Others argue that so few people use transit to get to work, and even
fewer for other reasons, that major new investments in transit capacity, except in a
limited number of situations, are not likely to reduce highway congestion
appreciably, if at all.44
The problem and empirical measurement of induced demand are a central
element in many of the debates about road traffic congestion. The theory of induced
demand suggests that building more road capacity will not solve road traffic
congestion because it merely “induces” travelers using other modes, driving on other
routes, or driving at other times of the day to travel on the new facility during the
peak period, resulting in congestion as bad as that suffered before the expansion.45
Some suggest it is even possible for congestion to become worse in the long run
after a road is built or expanded because the new capacity encourages more
development, resulting in proportionally more drivers than the new capacity added.46
Attaining a definitive answer to this question is difficult because of the confounding
factors of regional trends in population and employment growth and other things that
lead to changes in transportation habits.47 However, several studies show that
although induced demand is real, it typically takes a number of years for the new
capacity to be absorbed, suggesting that new capacity can reduce congestion in the
medium term.48 Moreover, other experts note that while congestion may reassert

43 Surface Transportation Policy Project, Easing the Burden: A Companion Analysis of the
Texas Transportation Institute’s 2001 Urban Mobility Study (Washington, DC, May 2001),
at [].
44 Wendell Cox and Randal O’Toole, “The Contribution of Highways and Transit to
Congestion Relief: A Realistic View,” Heritage Foundation Backgrounder, no. 1721
(January 27, 2004).
45 Pickerell, Don, “Induced Demand: Definition, Measurement and Significance,” in
Working Together to Address Induced Demand (Washington, DC: Eno Transportation
Foundation, 2002).
46 For a discussion of this issue see Downs, 2006, p. 104.
47 Pickerell, 2002.
48 Cervero, Robert, “Are Induced-Travel Studies Inducing Bad Investments?,” Access, no.

itself after the addition of major new capacity, the new facilities still serve more
travelers than before even if service quality is poor, and the increase in travelers on
the new or larger facility may take pressure off other facilities, improving travel over
the whole network.49
Congestion Pricing and Other Alternative Ways to
Ration Resources
Many economists argue that transportation congestion is caused by the way in
which service is rationed. In highway transportation, for example, because the
marginal cost of driving is so low, congestion is the main method for rationing peak-
period roadway space. Peak-period roadway space is in great demand for deep-
seated reasons that have to do with the need for face-to-face interaction in economic
and social situations. Thus, at certain times and in certain places, demand for
roadway space exceeds supply and vehicles have to queue for the next available
space to open up. It is argued that road traffic congestion could be reduced by using
different rationing methods. One approach is to limit roadway space to certain types
of vehicles or vehicles carrying a certain number of passengers, such as buses or
high-occupancy vehicle (HOV) lanes. Another method is to ban a vehicle or driver
from driving at certain times for one or more days a week. The method generally
favored by economists, however, is to use some sort of pricing mechanism, known
as congestion pricing or value pricing. Its supporters argue that not only does road
pricing have the potential for solving congestion, it also promotes the most efficient
use of highway infrastructure.
Detractors argue that road pricing unfairly favors higher-income drivers, may
cause severe mobility problems where no reasonable alternative exists, and may, if
it raises the cost of traveling in the most dense urban areas, lead to more sprawl and
highway congestion farther out from the urban core. Another argument against
tolling in general, of which congestion pricing is one form, is that drivers have often
already paid for the infrastructure and its maintenance through taxes and fees, and so
it amounts to a form of double taxation. Consequently, some suggest that such
strategies should be used only to fund and manage new capacity or should not be
used at all.
Demand for transit service in large cities is typically more concentrated, both
in time and by direction, than demand for highway travel. The result can be vehicle
overcrowding, service denial, and, because overcrowding tends to increase vehicle
dwell times (i.e., time spent at a station or bus stop to discharge and pick-up
passengers), overall slower speeds. Despite this, most transit agencies do not
differentiate fares on the basis of peak/off-peak service but instead have flat-fare

48 (...continued)

22, 2003, pp. 22-27, at [

2004%20-%20Induced%20T rave l%20Studies.pdf].

49 Downs, 2006; Poole, Robert, “New Evidence Questions the Reality of ‘Induced
Demand,’” Surface Transportation Innovations, no. 39 (January 2007), at
[ ht t p: / / www.r eason.or g/ sur f acet r a nspor t a t i on39.sht ml ] .

structures and offer unlimited ride passes.50 As is often pointed out, higher peak-
period fares would help to cover the higher costs of providing peak-period service
and might persuade some travelers to travel during less busy periods. Even where
higher peak-period fares are employed, however, they are not usually high enough
to substantially reduce demand peaking. Proposals to introduce differentiated fare
schemes to reduce overcrowding — or in places that have them to raise fares even
higher at congested times or places — are often viewed skeptically as a way for a
transit agency to generate more revenue, particularly from transit-dependent travelers.
Others fear such schemes might push public transit users to drive instead, causing
greater highway congestion.
In freight rail transportation, prices (or “rates” as they are more commonly
known) are already the main mechanism used to manage supply and demand. Rates
reflect the cost of providing freight rail service and demand. Demand for rail service
is largely a function of the overall strength of the economy and the ability of rail
transportation to compete with other modes, particularly trucks and barges. With
strong demand and constrained supply, economic theory would suggest, all else
equal, that rates will increase, providing greater resources for investing in expanding
supply. Although the situation is complex, because not all else is equal, the evidence
suggests that with greatly improved productivity and strong demand, the financial
health of the railroad industry has improved substantially since deregulation. This
has allowed railroad companies to make significant investments to maintain the
current system and to increase capacity in some places.51 Nevertheless, there is
widespread concern that the railroads will not be able to make sufficient investments
to keep up with demand.52
A number of reasons have been posited for the inability of railroads to invest
sufficiently in new capacity to keep up with demand. Clearly, expanding capacity is
a slow process, meaning it may take decades for supply and demand to find an
equilibrium, if it ever does. Moreover, in many congested urban areas, railroads find
it difficult to acquire land for new capacity.53 Port areas that could benefit from new
rail lines and terminal facilities are notoriously space-constrained. The railroads
argue that they suffer several inequities that hinder their ability to finance new
capacity. The railroads note that, unlike trucking and barge firms, they provide their
own infrastructure and must bear the long-term risks associated with owning fixed
assets. Furthermore, they argue, other modes pay less in taxes and fees than their use

50 Transportation Research Board, Fare Policies, Structures, and Technologies: Update,
Transit Cooperative Research Program (TCHRP) Report 94 (Washington, DC, 2003), table

2-6, at [].

51 CBO, 2006.
52 U.S. Government Accountability Office (GAO), Freight Railroads: Industry Health Has
Improved, but Concerns about Competition and Capacity Should Be Addressed, GAO-07-94
(Washington, DC, October 2006), at []; AASHTO,


53 Bryan, Joseph, Glen Weisbrod, and Carl Martland, Assessing Rail Freight Solutions to
Roadway Congestion: Final Report, NCHRP Project 8-42 (Washington, DC: Transportation
Research Board, October 2006), at [


of public infrastructure would warrant, putting the railroads at a competitive
disadvantage. Railroads also argue that they are subject to several industry-specific
laws that raise their costs in comparison with their competitors. These laws include
the Railroad Unemployment Insurance System and some remnants of the Interstate
Commerce Act.54 Ultimately, the railroads argue that despite improvements in their
financial situation since deregulation, they continue to have problems earning enough
to cover the cost of capital, hindering their ability to compete for financing in capital
market s.55
In this context, a number of public policy alternatives have been suggested to
alter the current rationing of public and private resources. One controversial proposal
is to impose greater taxes and fees on truck and barge companies to “level the playing
field” with railroads. Another is to provide government assistance to railroads to
mitigate some of the risks they face, with the goal of increasing the level of
investment and accelerating its current pace.56 On the other hand, some contend that
the railroads ought to make a greater financial commitment to solving problems
where they impose high external costs, such as places where rail operations
contribute significantly to highway congestion. For example, in 2002, northeastern
Illinois was estimated to have about 1,700 highway-rail grade crossings that caused
nearly 11,000 hours of motorist delay on a typical weekday.57 Contributions by the
railroads to highway-rail grade crossing improvements, such as grade separation
projects, however, tend to be a relatively small share of the overall cost.
A final consideration in the rationing of resources is what might be called the
costs of debate, review, and approval. Some argue that the costs of complying with
federal, state, and local regulation stemming from the multitude of planning,
environmental, and community involvement laws have substantially increased project
costs since the 1960s. These costs include the direct compliance costs of staff time
and the indirect costs of project delay that results in foregone opportunities in terms
of improved mobility, safety, and the like. Most agree that these laws serve an
important purpose and have several benefits. Nevertheless, many would like to
reduce the delay caused by the unnecessary duplication of effort and coordination
problems among the different parties.58

54 CBO, 2006.
55 Ibid., p.20.
56 Ibid.
57 Illinois Commerce Commission, “Motorist Delay at Public Highway-Rail Grade Crossings
in Northeastern Illinois,” Working Paper 2002-03 (July 2002), at
[ docs/rr/021114rrdelay.pdf].
58 See, for a general overview, U.S. Department of Transportation, Federal Highway
Administration, “Evaluating the Performance of Environmental Streamlining: Development
of a NEPA baseline for Measuring Continuous Performance” (Washington, DC), at
[ ht t p: / / www.envi r onment .f .gov/ s t r ml ng/ basel i ne/ i ndex.asp] .

Brief History of Transportation Congestion
Highway Transportation
In the early years of the century, before the mass production of motor vehicles,
congestion generally referred to overcrowded trolley lines and trolley cars in major
cities and downtown streets filled with pedestrians and horse-drawn passenger and
goods vehicles.59 For most of the 20th century, however, transportation congestion
meant road traffic congestion. The rapid rise of motor vehicle ownership,
particularly with the introduction of Ford’s Model T in 1908, together with
rudimentary road and traffic control systems, made urban road traffic congestion a
major transportation problem by the 1920s.60 Federal, state, and local governments
responded with a significant road-building effort in this period, although road traffic
congestion was largely “solved” by the Great Depression and the Second World War.
During the Second World War, with the massive diversion of resources to the
war effort, automobile use was widely discouraged. Public transit ridership boomed
again during this period, reaching an all-time high in the United States in 1946 of
23.4 billion trips.61 However, car ownership and motor vehicle travel rose rapidly
after the war causing another bout of concern with road traffic congestion,
particularly in and around cities.62 Congestion and the threat of future congestion
were among the reasons cited by President Eisenhower in his push to create the
Interstate Highway Program,63 although he was against the idea of urban interstates,
preferring instead bypasses that would allow through traffic to avoid the central
cities. Nevertheless, the cities themselves were insistent that urban interstates were
needed to solve urban congestion problems, and Congress obliged in the Federal-Aid
Highway Act of 1956 and the Highway Revenue Act of 1956 (P.L. 84-627).64

59 Wachs, M., “Fighting Traffic Congestion with Information Technology,” Issues in Science
and Technology (Fall 2002), at [].
60 Motor vehicle ownership increased from approximately 8,000 in 1900 to 27 million in
1930. See U.S. Department of Transportation, Federal Highway Administration, Highway
Statistics, Summary to 1985 (Washington, DC, 1987), p. 25.
61 American Public Transportation Association, 2006 Public Transportation Fact Book
(Washington, DC, 2006), at [].
62 Weingroff, Richard F., “The Genie in the Bottle: The Interstate System and Urban
Problems, 1939-1957,” Public Roads, vol. 64, no. 2 (September/October 2000), pp. 2-15,
at [].
63 In a speech Eisenhower argued, “The country urgently needs a modernized interstate
highway system to relieve existing congestion, to provide for the expected growth of motor
vehicle traffic, to strengthen the Nation’s defenses, to reduce the toll of human life exacted
each year in highway accidents, and to promote economic development.” Quoted in
Weingroff, R.D., “Original Intent: Purpose of the Interstate System: 1954-56,” at
[], as of December 28, 2006.
64 Schwartz, Gary T., “Urban Freeways and the Interstate System,” Southern California Law
Review, vol. 49 (1976), pp. 406-513.

Road capacity expanded rapidly following the passage of the 1956 acts that also
created the Highway Trust Fund. Less than 20 years later, by the end of 1974, about
36,000 miles of the 42,500 mile system were complete, with another 2,800 miles
under construction.65 Together with the improvement of other urban and rural road
networks, road capacity (measured by paved centerline miles of highways and
streets66) grew at about the same rate as motor vehicle travel from the mid-1940s to
the mid-1960s (Figure 1). The problem of road traffic congestion never disappeared
in major cities, but in the 1970s, the most vexing highway transportation problems
were energy, air quality and other environmental issues, and highway safety.
Figure 1. Motor Vehicle Travel and Road Capacity, 1941-2005

Source: U.S. Department of Transportation, Federal Highway Administration, Highway Statistics
(Washington, DC, various years).
Note: Paved centerline miles is the length of roads paved with some type of bituminous, Portland
cement concrete, or brick surface as measured along the center in one direction.
The growth in road capacity and motor vehicle travel began to diverge in a
major way during the 1970s, as shown in Figure 1. Except for slight dips associated
with the oil shocks of 1974 and 1979, motor vehicle travel continued to grow apace.
At the same time, growth in road capacity slowed as the interstate system neared
completion, maintenance requirements began to absorb more resources, and building
new capacity became more expensive and time-consuming as a result of new
environmental laws. Consequently, road traffic congestion began to climb quickly
again in the 1980s and has continued to rise ever since.
65 U.S. Department of Transportation, 1976, p. 481.
66 Paved centerline miles is the length of roads paved with some type of bituminous,
Portland cement concrete, or brick surface, as measured along the center in one direction.
As such, this metric does not account for the capacity provided by highways with more than
one lane in each direction. FHWA did not begin publishing lane-mile data until 1980.

Public Transportation
Public transportation congestion has not been a major issue since the end of the
Second World War, when transit ridership was at an all-time high. On the contrary,
the major issue, particularly through the 1950s and 1960s, was the overall lack of
riders resulting from increases in motor vehicle ownership, suburbanization, and
other changes in work and leisure.67 By the early 1970s, transit ridership was only
a quarter of what it had been at its peak in 1946, dropping from a high of 23.4 billion
trips to a low of 6.5 billion trips. In response, many streetcar systems were
abandoned in favor of diesel buses, and privately owned and operated transit systems
were taken over by public authorities. Public transportation has undergone
something of a resurgence since the mid-1970s with the building of a number of new
rail systems, particularly light rail, but also heavy rail and commuter rail. Since then,
transit ridership has increased modestly to about 10 billion trips in 2005.68 To put
this in context, however, the proportion of all trips made on transit declined by half
between 1969 and 2001, as trips by other modes, particularly in personal motor
vehicles, increased to a much greater extent.69
Although not as widespread as road traffic congestion, peak-period transit
overcrowding has become an issue in some cities with large numbers of transit
commuters and heavily congested roads and railways, such as New York; Chicago;
San Francisco; Washington, DC; and Boston. Peak-period overcrowding on the
subway in Washington, DC, for instance, has led to proposals for substantially higher
fares at the most heavily used times and stations.70 In addition, because most transit
buses do not run on roads with controlled access (e.g., high-occupancy vehicle
[HOV] and bus lanes), road traffic congestion also affects bus riders.
Freight Transportation
Until relatively recently, congestion has not been a major issue in freight
transportation. The building of the interstates, together with the existing rail, water,
and pipeline systems, provided adequate surface freight capacity from the 1960s
through the 1980s. According to many analysts, the biggest problem at this time was
antiquated federal regulation from laws dating to the late 19th and early 20th centuries.
Administered mainly by the now defunct Interstate Commerce Commission (ICC),
these regulations controlled prices and competition, leading to some major
inefficiencies in the transportation of goods. Deregulation beginning in the late

67 Smerk, George M., The Federal Role In Urban Mass Transportation (Bloomington, IN:
Indiana University Press, 1991).
68 American Public Transportation Association, “Transit Ridership Report, Fourth Quarter

2005,” April 4, 2006, at [


69 Polzin, Steven, and Xuehao Chu, Public Transit in America: Results from the 2001
National Household Travel Survey (Washington, DC, September 2005), at
70 Sun, Lena H., “Rush Hour Metro Fares May Rise As Much as $2.10,” The Washington
Post, December 14, 2006, p. A1.

1970s sparked a major reorganization within and across modes that overall has
provided shippers with cheaper, more efficient freight transportation and much
greater choice.
In railroading, for instance, federal regulation made it difficult to abandon little-
used or unprofitable lines. Thus, although railroad mileage peaked as early as 1916,
it changed little for the next 60 years. Overcapacity was a significant contributor to
the financial difficulties of the railroads that reached crisis proportions in the 1970s
and subsequently led to deregulation of the industry through the Staggers Rail Act
of 1980 (P.L. 96-448). The Staggers Act made it much easier for major railroads to
abandon lines or to sell or lease them to non-Class I railroads.71 Since then, the miles
of track owned and operated by Class I railroads have dropped precipitously from
271,000 in 1980 to 162,000 in 2006.72 Non-Class I railroad mileage consequently
has grown, although modestly. Despite less track, railroads today are able to move
more freight because technological changes allow them to run heavier, longer, and
faster trains. Indeed, freight rail ton-miles increased by 93% between 1980 and
2006.73 This has also been accomplished with relatively fewer locomotives, freight
cars, and employees, marking huge productivity gains since deregulation.
Deregulation also played a major role in the reorganization and growth of the
trucking industry. New laws such as the Motor Carrier Act of 1980 (P.L. 96-296)
and other changes freed up trucking companies to more directly compete against each
other, allowed the entry of new firms, and encouraged the development of efficient
truck operation and routing. The results have been generally lower prices and higher-
quality and more reliable service.
Among other things, deregulation played an important role in the shift toward
what has been called “coordinated logistics,” defined as “the integration of distinct
logistics activities, such as cross-modal coordination or the bundling of transportation
and inventory control.”74 Deregulation helped remove many of the modal and
jurisdictional barriers between carriers. Moreover, with industry consolidation and
improvements in productivity and profitability, carriers were able to introduce new
technologies and develop innovative services. For instance, over the past few
decades, trucking and railroad companies have created networks of trailer-on-flatcar
service that combine the advantages of rail and truck transportation. With cheaper
and more timely deliveries of goods, shippers have been able to save production and
distribution costs by developing longer and more complex supply chains and by
cutting back on their inventories of goods. Coordinated logistics, therefore, has

71 The Surface Transportation Board, the federal agency responsible for economic regulation
of the railroad industry, classifies freight railroads based on operating revenue. In 2006, the
classification was as follows: Class I, $346.8 million or more; Class II, $27.8 million to
$346.7 million; Class III, less than $27.7 million.
72 Association of American Railroads, Railroad Facts 2007 (Washington, DC), p. 45.
73 Ibid., p. 27.
74 U.S. Department of Transportation, Federal Highway Administration, “Freight Carriers:
From Modal Fragmentation to Coordinated Logistics,” undated white paper, p. 1, at
[ eme_papers/final_thm5_v4.htm] .

raised the importance of transportation in the logistics process and has placed greater
emphasis on seamless networks of multiple transportation modes.
Coordinated logistics has also been spurred on by extraordinary growth in
foreign trade. Foreign trade as a percentage of U.S. Gross Domestic Product (GDP)
has grown from 11% in 1970 to 26% in 2005.75 Consequently, the amount of goods
moving through foreign trade gateways — ports, border crossings, and airports —
has skyrocketed. For instance, waterborne merchandise trade almost tripled between
1970 and 2006, from 581 to 1,565 million tons.76 This growth has placed great
pressure on the gateways themselves, but also on the transportation networks that
serve them — primarily roads and rail lines — and the connection between modes.
Among other problems, most of these gateways are located in large urban centers that
suffer from high levels of road traffic congestion and have limited space for facility
expansion. Many experts now believe the efficiency gains resulting from
deregulation and other changes have largely run their course.77 After declining for
years, the cost of logistics to U.S. businesses appears to be increasing, partly because
of congestion (see Figure 2). In railroading, many lines and terminals are running
at or near full capacity. With little or no slack in the system, railroads have become
more susceptible to disruptive incidents, such as late loadings and unloadings,
breakdowns, and poor weather. Another problem as rail lines reach capacity is the
growing conflict between freight and passenger trains (Amtrak and commuter) that,
for the most part, use the same lines. As a result, delays are multiplying for both
freight and passenger trains, particularly in major urban areas that generate a lot of
freight and passenger traffic. In trucking, productivity is now largely dependent on
road congestion, the supply of qualified truck drivers, and fuel costs.

75 U.S. Bureau of Economic Analysis, National Income and Product Accounts, at
[] .
76 U.S. Army Corps of Engineers, Waterborne Commerce Statistics of the United States
2005, National Summaries (New Orleans, LA, 2007), p. 1-3, at
[]; U.S. Army Corps of
Engineers, 2006 Preliminary Waterborne Commerce Statistics: National Totals and
Selected Inland Waterways (New Orleans, LA, October 23, 2007), p. 1, at
77 U.S. Department of Transportation, Federal Highway Administration, “Regulation: From
Economic Deregulation to Safety Regulation,” undated white paper, at
[ eme_papers/final_thm8_v4.htm] .

Figure 2. U.S. Cost of Logistics, 1984-2005
(percentage of Gross Domestic Product)

Ye a r
Source: Council of Logistics Management, State of Logistics Report (Washington, DC, 2006).
Legislative History of Transportation Congestion
In line with the rapid growth of motor vehicle ownership and travel, federal
surface transportation policy for most of the 20th century focused on road connectivity
and capacity, particularly with a view to providing basic access in rural areas and
then intercounty and interstate roads. Urban road traffic congestion warranted a
certain amount of attention in the early Federal-Aid Highway Acts, including the
Federal-Aid Highway Act of 1956. Federal transit funding, beginning in the 1960s,
was also partly predicated on the argument that it would relieve road traffic
congestion.78 As the interstate building program neared completion in the 1980s and
road traffic congestion was growing apace, federal policy makers began a
fundamental reassessment of surface transportation policy. The result was the
Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA), P.L. 102-240.
Intermodal Surface Transportation Efficiency Act of 1991
(P.L. 102-240)
In the deliberations of the congressional committees that culminated in the
passage of Intermodal Surface Transportation Efficiency Act (ISTEA), there was
recognition that urban road traffic congestion was a major problem.79 Unlike in the
past, however, some viewed road capacity building as a flawed strategy for dealing
78 Smerk, 1991.
79 U.S. Congress, Senate Committee on Environment and Public Works, S.Rept. 102-71,
June 4, 1991; U.S. Congress, House Committee on Public Works and Transportation,
H.Rept. 102-171(I), July 26, 1991.

with the issue. This view was summed up by Senator Daniel Moynihan in the
introductory statement of the Senate report. Talking about the building of the
interstate system, he argued the following:
[T]he plain fact is that traffic congestion has grown during this period of massive
highway construction. We have to face the fact that even if we had greater
resources than we do, adding to highway capacity does not any longer seem a80
promising road to increased highway efficiency.
Congressional leaders also expressed the concern that solutions to transportation
problems that encouraged more driving would lead to more air pollution, thereby
undermining the provisions of the recently enacted Clean Air Act Amendments of
1990 (CAAA), P.L. 101-549. A third major concern was that scarce resources should
be used first and foremost to maintain and improve the current highway system over
system expansion.
Rather than design a new road-building program, leaders in both the House and
the Senate sought to fashion a program to enhance the efficiency of a transportation
system that was largely in place. This new program would be based on highway
system maintenance; more transit funding; greater funding flexibility; intermodalism;
enhanced state and metropolitan planning; improved operations, including
development and deployment of advanced technologies (e.g., technologies to
improve roadway monitoring, enhance traveler information, and enable the electronic
payment of tolls); and efforts to improve safety, energy efficiency, and pollution
control. The new surface transportation bill also required the designation of a new
National Highway System (NHS) to prioritize federal help for the most heavily
traveled routes of the Interstate Highway System, the Strategic Highway Network,81
and Federal-Aid Primary System.
A fundamental theme in the development of ISTEA was that states and localities
should be free to fashion their own solutions to local problems, a tenet that became
known as “flexibility.” While recognizing that congestion was a problem, the
committees understood that it was not a problem everywhere, hence the need for
flexibility. In this regard, the House Public Works and Transportation Committee
report noted, “The new system reflects the Committee’s recognition of the need to
relieve congestion in urban and suburban America, while at the same time addressing
the mobility and access needs of Rural America.”82
In reworking the surface programs, the large Surface Transportation Program
(STP), authorized at $24 billion over the life of the bill, was at the core of the

80 S.Rept. 102-71, June 4, 1991, p. 4.
81 At this time, the Interstate Highway System was approximately 45,000 miles in length.
The interstates were part of the Federal-Aid Primary System, which also included
approximately 260,000 miles of mostly rural arterials and some urban principal arterials.
The Strategic Highway Network (STRAHNET) was, and still is, a system of roadways
identified as being important for national defense. In addition to the interstates, the
STRAHNET at this time included another 15,000 miles of non-interstate roads.
82 U.S. Congress, H.Rept. 102-171(I), July 26, 1991, p. 6.

flexibility provisions. STP funds were made available for highway capital projects
but could be “flexed” to transit if desired and if certain other conditions were met.
As the House Committee noted, “For those with congested urban areas, flexibility
may mean more transit solutions, while for rural areas or those experiencing
economic growth, flexibility may mean more highways.”83 Within certain
parameters, flexibility was also provided for switching funds between the different
parts of the highway system, as projects could be on any part of the system except
local and rural minor collectors. Moreover, STP funds could be used for a bridge
project on any public road, not just those on the federal-aid system.
ISTEA also authorized a substantial increase in federal transit funding over
previous authorizations, nearly $32 billion over the life of the bill. Many expected
that additional funding would be devoted to transit from flexed STP funds and
another new program, the $6 billion Congestion Mitigation and Air Quality program
(CMAQ), which was authorized to provide new funds for projects to help states and
localities meet the requirements of the Clean Air Act Amendments (CAAA) of 1990.
Funding was aimed primarily at reducing pollutants emitted by reducing motor
vehicle travel, particularly single-occupant vehicle travel. Because the most polluted
places tend to have the worst road traffic congestion, it was believed that many
projects funded under CMAQ to reduce pollution would reduce road traffic
congestion as well. However, CMAQ prohibited spending on more traditional
congestion relief projects, such as new road capacity that would be primarily used by
single-occupant drivers. In addition, building new capacity could violate the
requirements in the Clean Air Act and ISTEA that state and metropolitan plans
“conform” to the emissions levels set forth in the air quality State Implementation
Plan (SIP) as required by CAAA. A 10-year assessment of the program found that
about 44% of CMAQ funds were spent on transit projects and another 33% on traffic
flow improvement projects such as incident management, HOV lanes, and traffic
signal improvements.84
ISTEA also advanced a few other congestion-related programs that were federal
program innovations. First was the idea of intermodalism — planning and financing
projects that enhance the links between modes. In this regard, states and
metropolitan areas were required to consider the transportation systems as whole in
the planning process and to include participation from all stakeholders, including the
freight community. Funds were also made available for highway projects to
accommodate other transportation modes and for carpool projects, such as fringe and
corridor parking facilities and programs, and bicycle transportation and pedestrian
walkways. Second, ISTEA placed more emphasis on funding highway operations,
including the establishment of a new program to fund the development and
deployment of advanced technology in transportation, known as the Intelligent
Vehicle/Highway Systems Program (IVHS). Now known as Intelligent
Transportation Systems (ITS), the program was originally authorized with $660
million over the six year life of the act. Third, to enhance the ability of metropolitan
areas to coordinate and fund the development of their transportation systems, ISTEA

83 Ibid., p. 7.
84 Transportation Research Board, The Congestion Mitigation and Air Quality Improvement
Program: Assessing 10 Years of Experience, Special Report 264 (Washington, DC, 2002).

increased the responsibilities of metropolitan planning organization (MPOs) and
required the development of congestion management systems at both the
metropolitan and state level. The requirement for a congestion management system
at the state level was subsequently dropped in the National Highway System
Designation Act of 1995 (P.L. 104-59).85 Fourth, ISTEA provided funding for up to
five projects in the Congestion Pricing Pilot Program and allowed greater use of
federal funds on toll roads than in the past.
National Highway System Designation Act of 1995
(P.L. 104-59)
ISTEA required the designation of a new category of highways, the National
Highway System (NHS), to be worked out in consultations between the USDOT and
the states. The designation of the 155,000-mile NHS system was the primary
purpose of the NHS Act. However, the NHS Act included several other provisions
amending the federal programs, some with relevance to the issue of mobility and
congestion. Among them were the authorization of two new financing mechanisms:
the State Infrastructure Bank (SIB) pilot program and what became known as Grant
Anticipation Revenue Vehicle (GARVEE) bonds. The SIB pilot project allowed a
handful of states to use some of their highway and transit funds to capitalize a
revolving fund. The GARVEE bonds were developed from Section 311 of the NHS
Act that expanded the use of federal-aid highway funds for bond financing. A
number of intermodal projects, including the Alameda Corridor project, were
advanced because of these new provisions.
Transportation Equity Act for the 21st Century
(P.L. 105-178; P.L. 105-206)
The Transportation Equity Act for the 21st Century (TEA-21), as amended (P.L.
105-178; P.L. 105-206), enacted June 9, 1998, maintained the essential structure of
the programs created in ISTEA with an increase in funding (in nominal terms) of
40%. Of the total $218 billion authorized, $177 billion was allocated for highways
and $41 billion for transit, although TEA-21 continued and enhanced the flexing of
monies between modes as introduced by ISTEA in 1991.86
Several programs begun in ISTEA were retained and expanded under TEA-21.
CMAQ was retained with more funding ($8.1 billion) and expanded eligibility
criteria. ITS funding was raised to $1.282 billion, and a new ITS program, the
Commercial Vehicle Information Systems and Networks (CVISN) Program, was
established and funded at $184 million. With the ultimate goal of improving the
efficiency and safety of commercial motor vehicle operations, the CVISN program
was created to make use of information systems and communications networks by
developing industry standards and demonstrating potential benefits. Three areas

85 U.S. Government Accounting Office, Transportation Infrastructure: States’
Implementation of Transportation Management Systems, GAO-RCED-97-32 (Washington,
DC, 1997), at [].
86 U.S. Department of Transportation, Transportation Equity Act for the 21st Century — A
Summary (Washington, DC, 1998).

were initially targeted under the CVISN program: safety information exchange,
credentials administration, and electronic screening.87 The Congestion Pricing Pilot
Program was renamed the Value Pricing Pilot Program and funded at a higher,
though still very modest, level ($51 million).
TEA-21 also created a few new programs. Some of these came under the
banner of innovative financing, including the Transportation Infrastructure Finance
and Innovation Act (TIFIA) and the Railroad Rehabilitation and Improvement
Financing (RRIF) program. TIFIA was to provide up to $10.6 billion in credit
assistance to large projects of national significance (generally projects over $100
million). The RRIF program was set up to provide loan and loan guarantees up to
$3.5 billion, of which not less than $1 billion was to be available to non-Class I
railroads. Two new infrastructure grant programs — the National Corridor Planning
and Development Program and the Coordinated Border and Infrastructure Program
— were also created and jointly funded at $140 million per year for FY1999 through
FY2003.88 The first was conceived primarily as an economic development tool
(although congestion costs were one factor to be used in determining projects) and
the second was intended to alleviate congestion and improve mobility at the borders.
Since FY2000, nearly all the funds in this program have been earmarked in
appropriation bills.
Safe, Accountable, Flexible, Efficient Transportation Equity
Act — A Legacy for Users (P.L. 109-59)
After a number of hearings prior to reauthorization of TEA-21 in which
transportation congestion was a major focus, the initial legislative proposal from theth
House of Representatives (H.R. 3550) in the 108 Congress included a number of
new provisions in Subtitle B, entitled “Congestion Relief.” Two provisions were
seen as being particularly innovative. The first was the Motor Vehicle Congestion
Relief Program, which would require states with an urbanized area over 200,000 to
set aside apportioned funds under several existing programs to be spent on projects
that enhance capacity and relieve congestion. The proposed set-aside was 10% of a
state’s total apportionments multiplied by the percentage of the state’s population
in urbanized areas of 200,000 or more. The second innovative proposal was to fund
ITS technologies at a much higher level and to speed up their deployment. H.R. 3550
would have authorized about $4 billion during FY2004-FY2009, with about $3
billion of this amount for expedited deployment. This was up from about $230
million per year toward the end of TEA-21 (not including federal-aid highway funds89

allocated by the states to deploy ITS).
87 U.S. Department of Transportation, Federal Motor Carrier Safety Administration,
Introductory Guide to CVISN (Washington, DC, 2000), at [
88 ISTEA had identified 21 high priority corridors, and the NHS Designation Act had added
another 8 corridors. ISTEA provided funds for feasibility and design studies.
89 See CRS Report RL32226, Highway and Transit Program Reauthorization Legislation
in the 2nd Session of the 108th Congress, by John W. Fisher.

H.R. 3550 proposed a new $6.6 billion allocated program called Projects of
National and Regional Significance to fund important high-cost facilities ($500
million or more or greater than 75% of a state’s annual apportionment), including
freight rail projects eligible under Title 23 U.S.C. Also included in the bill was a
new Freight Intermodal Connectors program to be funded by formula at the level of
$1.37 billion over six years and a Freight Intermodal Distribution Pilot Grant
Program funded at $30 million over five years as a takedown from the Freight
Intermodal Connectors authorization.90 This latter program was intended to provide
grants to facilitate intermodal freight transportation initiatives at the state and local
levels to relieve congestion and improve safety, and to provide capital funding to
address infrastructure and freight distribution needs at inland ports and intermodal
freight facilities. As passed by the House, two tolling provisions were also included
in H.R. 3550, one to permit states to allow drivers to pay to use HOV facilities as part
of a variable toll-pricing program and the other to permit the construction of new
lanes on interstates to be funded by tolls.91
The reauthorization of the surface transportation programs was not passed in the
108th Congress but was eventually completed in the 109th Congress and signed into
law by the President on August 10, 2005. The Safe, Accountable, Flexible, Efficient
Transportation Equity Act — A Legacy for Users (SAFETEA) provides a general
increase in transportation funding with a six-year total of $286.4 billion for programs
from FY2004 through FY2009. This represents a 31% increase in nominal terms
over the $218 billion provided over the six years of TEA-21 (FY1998-FY2003).92
As enacted, SAFETEA largely retains the structure of the surface transportation
programs begun under ISTEA, with a large proportion of funding going to the
established “core” highway programs (such as the Surface Transportation Program,
the National Highway System, the Interstate Maintenance Program, and the Bridge
Program) and public transportation.93 The Congestion Relief subtitle of SAFETEA
contains just one program, the new Real-Time System Management Information
Program. This program, with no separate funds of its own, is designed to encourage
states to develop a real-time traffic information system to improve highway
operations and reduce congestion. The rest of the Congestion Relief programs, as
proposed in H.R. 3550, were either shifted elsewhere in the act or deleted. ITS
funding was not retained as a separate program but was “mainstreamed” as an
eligible category in the core programs. CMAQ continues at a higher funding level,
and project eligibility is expanded to include projects that might have a more direct
impact on congestion. Table 1 shows the authorization levels of SAFETEA’s titles
and some selected programs for FY2005 through FY2009.

90 U.S. Congress, House Committee on Transportation and Infrastructure, H.Rept. 108-452,
March 29, 2004.
91 A similar provision was included in S. 1072.
92 CRS Report RL33119, Safe, Accountable, Flexible, Efficient Transportation Equity Act
— A Legacy for Users: Selected Major Provisions, coordinated by John W. Fisher.
93 Ibid.

Table 1. SAFETEA Authorization Levels, by Legislative Titles
and Selected Programs, FY2005-FY2009
(in millions of dollars)
Total Authorization
Selected SAFETEA Title/ProgramFY2005-FY2009
Title I Federal Aid Highways199,490.476
Interstate Maintenance Program25,201.595
National Highway System30,541.833
Bridge Program21,607.422
Surface Transportation Program32,549.757
Congestion Mitigation & Air Quality Improvement Program (CMAQ)8,609.100
National Corridor Infrastructure Improvement Program1,948.000
Coordinated Border Infrastructure Program833.000
Projects of National & Regional Significance1,779.000
National Corridor Planning & Development & Coordinated Border140.000
Infrastructure Programs
Freight Intermodal Distribution Pilot Grant Program30.000
Value Pricing Pilot Program59.000
Title II Highway Safety3,131.592
Title III — Public Transportation45,313.000
Title IV Motor Carrier Safety 2,519.829
Titles V-X (excluding rescission of unobligated balances of5,003.940
highway contract authority in Title X)
Source: CRS Report RL33119, Safe, Accountable, Flexible, Efficient Transportation Equity Act
A Legacy for Users: Selected Major Provisions, coordinated by John W. Fisher.
SAFETEA does provide states with slightly more latitude in using tolls to build
or expand interstate capacity and to improve operational efficiency to reduce
congestion. The Value Pricing Pilot Program was reauthorized at a higher level: $11
million for FY2005 and $12 million annually for FY2006-FY2009. In addition,
SAFETEA includes provisions for a limited number of pilot projects to test the
viability of the use of tolling on existing facilities including HOV facilities and for
tolling to fund new interstate capacity.
SAFETEA also created the new Projects of National or Regional Significance
program, but with funding set at $1.779 billion for FY2005 through FY2009, not
$6.6 billion as proposed in H.R. 3550, and all the funds earmarked in the act. The
new Freight Intermodal Connectors program was dropped before final passage of the
bill, but the Freight Intermodal Distribution Pilot Program remained with $30 million
authorized through FY2009. Again, this $30 million was earmarked in the bill.
SAFETEA also reauthorized the Coordinated Border Infrastructure Program as a new
apportioned program, with funding set at $833 million from FY2005 though FY2009.
Existing innovative funding provisions were extended and modified to some
degree in SAFETEA. For instance, the minimum project size for TIFIA projects was
reduced from $100 million to $50 million for most projects and from $30 million to
$15 million for ITS projects. SAFETEA also allowed for broadened use of SIBs and

Private Activity bonds. The RRIF was expanded tenfold under SAFETEA, from $3.5
billion to $35 billion in loans. Of this, $7 billion is reserved for non-Class I railroads.
The legislation also added to the list of priorities in using such loans “enhancing rail
infrastructure capacity and alleviating rail bottlenecks.” SAFETEA also added a new
federal grant program for relocating rail track that interferes with motor vehicle
Transportation Congestion: Concepts, Measures,
and Trends
Transportation congestion exists when demand for a transportation facility or
vehicle is greater than its capacity and the excess demand causes a significant drop
in service quality, such as speed, cost, and comfort, depending on the mode and
specific situation. For example, when too many drivers compete for road space, the
result is usually a significant drop in traffic speed but also higher vehicle operating
costs and, with bumper-to-bumper, stop-and-go conditions, an increase in driver
stress. In freight railroad transportation, train speeds may suffer when demand begins
to reach capacity, and because shippers directly pay for access to rail infrastructure,
higher rates theoretically may be another indicator of congestion. Depending on the
situation, congestion in public transit may result in vehicle overcrowding — possibly
resulting in service denial and reduced passenger comfort — slower vehicle speeds,
and higher peak-period fares.
From the viewpoint of a multi-modal passenger trip or freight shipment, the
possibility for congestion exists not only within each mode but also in the
connections between modes. Poor or overstretched intermodal connections are
another part of the transportation system that may damage service quality. Moreover,
inefficient intermodal connections may cause problems within a mode as unexpected
delays interfere with other trips and shipments farther down the line. For example,
a delayed ship-to-truck transfer in a major metropolitan area may result in the truck
traveling during peak-period traffic.
Ideally, transportation congestion should be defined and measured from the
perspective of the end user — a traveler or a freight shipment. Congestion, therefore,
could be measured by the extent to which excess demand slows or otherwise harms
a passenger trip or freight shipment from the origin to the destination.94 In some
situations, such as the transportation of packages by an express carrier, such as UPS
and FedEx, it may be possible for the carrier to collect data and monitor movements
for business purposes. However, in most situations, for public policy purposes,
because measuring trips from origin to destination is difficult to accomplish in a large
scale and meaningful way, measures of congestion typically focus on service
problems within a mode. Moreover, within each mode, many measures of congestion
are limited to a specific transportation facility. This is especially the case in highway
transportation. For example, highway engineers typically refer to speed or level of

94 Giglio, Joseph M., Mobility: America’s Transportation Mess and How to Fix It
(Washington, DC: Hudson Institute, 2005).

service (LOS) on a particular road segment. Measurements on these segments are
then sometimes aggregated to develop systemwide measures of highway congestion.
Mode-specific and facility-specific measures of congestion are not wholly
satisfactory indicators of capacity problems in transportation service because they fail
to measure aggregate impacts across the whole system. On the other hand, some
transportation experts have noted that the focus on facility congestion instead of the
effect of congestion on passenger and freight trips may also overstate its importance.
For instance, freeway congestion may not be as bad as it seems if seen in the context
of an entire automobile commute trip, including the time it takes to park and walk to
the office.95 Similarly, it might be true that the effect of freight bottlenecks might not
be as bad as is generally believed if seen from the perspective of the entire supply
Whether facility-based or trip-based, another criticism of transportation-based
congestion measures is that they ignore the land-use context within which travel is
taking place. In transportation planning parlance, they measure mobility but not
accessibility. Accessibility explains the seeming paradox of why the most congested
places are also the most economically vibrant, even when the congestion is long
lived. Manhattan, for example, may be one of the most congested places on earth,
but it also provides access to an enormous number of opportunities in terms of
homes, jobs, retail outlets, restaurants, recreation, etc. A study of accessibility in
Minneapolis, MN, for example, found that while traffic congestion more than
doubled between 1990 and 2000 (measured in annual delay per person), access to
opportunities by car, in this case the number of jobs, increased more quickly.96 Seen
from this perspective, the performance of the transportation system, in concert with
land-use, actually improved in the 1990s rather than deteriorated, as congestion data
alone would suggest.
Unfortunately, as it stands today, national data do not exist to examine the
effects of congestion on accessibility as opposed to mobility. Nor do we have the
means to examine the effects of congestion on passenger trips and freight shipments
from end-to-end, including the efficiency of intermodal connections. The
transportation congestion measures employed in most instances, including in this
report, are both facility- and modally-based, with the inadequacies this entails.
Several measures of congestion, particularly in freight rail and public transit, are
gross indicators of capacity utilization using aggregate measures across the whole
system. Moreover, no measures of intermodal terminal congestion per se exist. The
measures of congestion presented here, nonetheless, represent the best available
information today using publicly available data.

95 Taylor, 2002.
96 El-Geneidy, Ahmed M. and David M. Levinson, May 2006.

Measures and Trends in Road Traffic Congestion
Efforts to define and measure road traffic congestion have increased over the
past few decades as congestion itself has grown.97 Still, congestion has proven
difficult to measure at the national level because of the size and diversity of the
highway system and because traffic problems can occur anywhere at any time of the
day or night for a number of different reasons. Moreover, what constitutes a
“congestion problem” is highly subjective. One frequently cited national road traffic
research effort is the Urban Mobility Program at the Texas Transportation Institute
(TTI). TTI defines traffic congestion as an excess of demand in relation to supply (or
capacity) such that travel speeds are slower than normal, where normal is defined as
free-flow speed. TTI derives travel speeds by relating the theoretical capacity of a
roadway segment to the average number and type of vehicles traveling the segment.
Speed estimates are then used to calculate travel delay. TTI uses data from FHWA’s
Highway Performance Monitoring System.98
Travel delay measures the extra time it takes to make a trip and can be expressed
in several different ways, such as total delay, delay per traveler, and as a travel time
index. The travel time index measures the ratio of travel time in the peak period to
travel time at free-flow conditions. Thus, a Travel Time Index of 1.35 indicates a 20-
minute free-flow trip takes 27 minutes in the peak-period.
In related research, TTI is developing measures of travel time reliability. Travel
time reliability measures the variability of travel times. When the highway system
is unreliable, travelers and shippers must build in extra time to avoid being late. TTI
measures travel time reliability via its Buffer Time Index (BTI). The BTI measures
the extra time needed to ensure that a traveler or freight shipment will arrive on time
according to a predetermined standard, typically 95% of trips. A BTI of 43%, for
instance, indicates that a traveler needs to add an extra 43% to the average travel time
of a trip to arrive on time 19 out of 20 times (95% of trips).99
Some suggest that reliability is more important to both travelers and shippers
than average delay. It seems reasonable to propose that most commuters would
prefer to spend an extra 5 minutes to and from work each day than to endure an
unexpected delay of 50 minutes on just one journey a week, a delay causing problems
with arriving at work on time or picking up a child from school or daycare. Similarly,
shippers often place greater value on being able to predict reliably when a shipment
will arrive than on the speed with which it got there. In some cases, such as just-in-
time manufacturing and distribution operations, shippers and carriers can face
penalties for making late or, in some cases, early deliveries.

97 Transportation Research Board, Quantifying Congestion, Volume 1, National Cooperative
Highway Research Program, Report 398 (Washington, DC, 1997).
98 Texas Transportation Institute, Urban Mobility Report 2007 (College Station, Texas),
Appendix A, at [].
99 Texas Transportation Institute and Cambridge Systematics, Monitoring Urban Freeways
in 2003, report prepared for the U.S. Department of Transportation, Federal Highway
Administration, December 2004, at [

In its annual Urban Mobility Report, TTI aggregates road segment estimates for
an entire urban area system of freeways and arterials. The same methodology has
been used by other researchers to identify and measure delay and, in some cases,
reliability at specific places, such as bottlenecks,100 truck bottlenecks,101 and border
crossings,102 as well as roads on the federally adopted National Highway System.103
The FHWA is using similar measures to examine congestion on major travel
corridors defined by Interstate routes, such as I-5 traversing California, Oregon, and
Washington. However, in this research program, FHWA is using data collected from
trucks themselves using Global Positioning System (GPS) technology.104
One of the main criticisms of TTI’s work on urban road traffic congestion is that
it does not directly measure congestion in any urban area, but relies instead on
estimates of congestion based on a number of theoretical relationships. For a time,
this meant that TTI was unable to account for improvements in speeds resulting from
operational improvements — such as freeway entrance ramp metering, incident
management programs, and traffic signal coordination programs — nor the effects
of public transit. TTI has since begun including these variables in its models, but the
overall criticism that its estimates of congestion are not direct empirical
measurements still stands.
Another major criticism has to do with the estimation of congestion by
comparing traffic speeds to free-flow conditions. A number of experts point out that
such models can never fully account for induced traffic and that, as problematic as
this may be theoretically, as a practical matter, eliminating congestion for all peak-
period travelers is wholly unrealistic because the costs would be overwhelming.
Thus, congestion-free peak-period travel in major metropolitan areas “is a purely
notional idea, not a conceivable description of the world we might choose to provide

100 American Highway Users Alliance, Unclogging America’s Arteries: Effective Relief for
Highway Bottlenecks, 1999-2004 (Washington, DC, February 2004), at
[http://www.highways .org/ pdfs/bottleneck2004.pdf].
101 Cambridge Systematics, “An Initial Assessment of Freight Bottlenecks on Highways,”
report prepared for U.S. Department of Transportation, Federal Highway Administration,
October 2005, at [].
102 Texas Transportation Institute and Battelle Memorial Institute, “International Border
Crossing Truck Travel Time for 2001,” report prepared for U.S. Department of
Transportation, Federal Highway Administration, April 2002, at [
103 U.S. Department of Transportation, Federal Highway Administration, Office of Freight
Management and Operation, The Freight Story: A National Perspective on Enhancing
Freight Transportation (Washington, DC, November 2002), p. 13, at
[ _analysis/freight_story/freight.pdf].
104 U.S. Department of Transportation, Federal Highway Administration, Office of Freight
Management and Operation, Freight Performance Measurement: Travel Time in Freight
Significant Corridors, FHWA-HOP-07-071, December 2006, at
[ h t t p : / / ops .f .gov/freight/freight_analysis/perform_meas/fpmtraveltime/traveltime

for.”105 Moreover, using free-flow speed in the calculation of congestion can lead to
some results that do not square with reality. For instance, if widening a road
improves the peak-period average speed but is accompanied by a proportionally
greater increase in the speed limit, the calculated amount of congestion will increase
after the improvement. In addition, a small change in average conditions, such as a
decrease of a few miles an hour, may appear to be a significant congestion problem
when measured over a large number of drivers.106
Empirical research on the relationship between freeway speed and vehicle flow
shows maximum vehicle throughput at something less than free-flow speed, about
50 miles an hour. This too brings into question a congestion calculation based on
free-flow speed. As Figure 3 shows, when there are few vehicles traveling on a
freeway segment, as might be the case very early in the morning, average speeds are
high, at about 60 miles per hour (mph), but overall throughput is low, at around 300
vehicles per lane per hour. As volumes build, vehicle throughput increases to around

1,800 vehicles per lane per hour and average speeds decline by about 10 to 15 mph.

At this point, as the number of vehicles coming onto the road continues to increase,
the volume of vehicles begins to overwhelm capacity and speeds decline
precipitously. As speeds decline in this instance, vehicle throughput declines.107
Overall, this line of criticism concludes that estimating congestion using the
unattainable ideal of free-flow conditions, and with it the costs of congestion (see
below), tends to overstate its impact on society. This and other criticisms
notwithstanding, the TTI estimates of urban road traffic congestion are widely used
because they provide the only national picture of road traffic congestion on an annual
basis and, hence, are useful for monitoring changes in congestion over time.
Nevertheless, figures purporting to quantify the billions of hours of time lost (and
their associated monetary value), numbers often used in newspaper headlines to
dramatize the problem, ought to be viewed somewhat skeptically.
A very important finding from the work by TTI and others is that both roadway
demand and roadway capacity are subject to short-term and long-term variations.
Demand varies by day of week, time of day, and season, and in response to planned
special events, such as professional football games, music festivals, and the like.
Most road traffic congestion occurs on weekday mornings and evenings because of
trips associated with jobs and school. Roadway capacity, on the other hand, is
defined by the type of facility (number of lanes, access, etc.), its condition, and by
events that may temporarily reduce capacity, such as traffic incidents, work zones,
weather, railroad crossings, toll facilities, and commercial truck pickup and delivery108

in urban areas.
105 Goodwin, Phil, “The Economic Costs of Road Traffic Congestion,” Discussion Paper,
Transport Studies Unit, University College London, 2004, p. 13, at [
106 Ibid.
107 Downs, 2006, Appendix A.
108 Oak Ridge National Laboratory, Temporary Losses of Highway Capacity and Impacts
on Performance: Phase 2 (Oak Ridge, TN, October 2004), at [

Figure 3. The Relationship Between Speed and Vehicle Flow
on Freeways

Source: Downs, Anthony, Still Stuck in Traffic, Brookings Institute Press (Washington, DC, 2006).
According to the current research, about 40% of urban road traffic congestion
is caused by capacity problems and another 5% is caused by poor signal timing
(Figure 4). About 55% of congestion is the result of a temporary loss of capacity,
with incidents (crashes, disabled vehicles, etc.) accounting for 25%, weather 15%,
work zones 10%, and other events 5%.109
Current Trends in Road Traffic Congestion. Most experts agree that
urban road traffic congestion has intensified and become more widespread during the
past quarter century. TTI data from 437 urban areas covering the period 1982
through 2005 indicate that total travel delay has increased five-fold and delay per110
peak-period traveler has nearly tripled. On average, delay increases with city size,
but delay in small urban areas (those with a population of less than 500,000) has
grown more quickly during this time period. Figure 5 demonstrates this in the 85
urban areas for which TTI provides detailed data. In addition, the morning and
evening rush periods have lengthened and a greater share of roadways are congested.
For instance, in the Louisville metropolitan area — a medium-sized urban area with
108 (...continued)
cta/Publications/tlc/tlc2_title.shtml ].
109 Cambridge Systematics and Texas Transportation Institute, “Traffic Congestion and
Reliability: Trends and Advanced Strategies for Congestion Mitigation,” report prepared for
U.S. Department of Transportation, Federal Highway Administration (September 1, 2005),
at [].
110 Texas Transportation Institute, 2007.

a population of about 900,000 that covers parts of Kentucky and Indiana — the share
of the road system congested has risen from 35% in 1982 to 52% in 2005. Moreover,
the number of “rush hours” has increased from 4.2 hours per day to 7.2 hours.
Figure 4. Proximate Causes of Road
Traffic Congestion

Source: Cambridge Systematic and Texas Transportation Institute, Traffic Congestion and
Reliability: Trends and Advanced Strategies for Congestion Mitigation, report prepared for U.S.
Department of Transportation, Federal Highway Administration (September 1, 2005).
Despite becoming more widespread, road traffic congestion is still heavily
concentrated in a few of America’s largest urban places. The 10 largest urban areas
by population account for nearly one-half of total delay, though only about one-
quarter of the U.S. population and the top 20 account for two-thirds of total delay and
one-third of the population. Los Angeles suffered the most delay in 2005, with 72
hours of annual delay per peak-period traveler and a Travel Time Index of 1.5.
Urban road traffic congestion has increased because motor vehicle travel has
grown rapidly, outstripping the existing road capacity and efforts to add new capacity
and improve throughput with operational treatments. In the 437 urban areas studied
by TTI, daily vehicle miles traveled on freeways grew by 128% between 1982 and
2005 and by 77% on arterials, while freeway and arterial lane-miles increased by only

41% and 37% respectively. Nationally, lane-miles grew by 4% and VMT by 87%

during this period.111
111 U.S. Department of Transportation, Federal Highway Administration, Highway Statistics
(Washington, DC, Annual Issues), at [].

Figure 5. Road Traffic Congestion, 1982-2005

erVery Large Urban Areas
d Tr
rio85 Urban Area Average
eakLarge Urban Areas
r P
30 pe
yMedium Urban Areas
20s ofSmall Urban Areas
0 198 2 198 3 1984 1985 1986 1987 1988 1989 1 990 1 991 19 92 199 3 1 994 1 995 1 996 19 97 199 8 199 9 2000 2001 2002 2003 2004 2005
Y ear
Source: Texas Transportation Institute, Urban Mobility Report 2007 (College Station, Texas, 2007).
Motor vehicle travel has grown rapidly for a number of reasons, including
substantial growth in population, jobs, and national income; increased vehicle
availability; and growth in metropolitan areas, particularly the suburbs. Between
1980 and 2005, the United States added 69 million people (a 30% increase), 42
million to the ranks of the employed (a 43% increase), 86 million motor vehicles (a

53% increase), and gross domestic product (GDP) grew by 113% in real terms.112

Both population and job growth have been concentrated in metropolitan areas, most
especially in low-density suburban rings that are difficult to serve with public transit.
A metropolitan suburb-to-suburb commute is today, by far, the most common type
of commute.113 As result, most people drive alone to work — 77% in 2005, up from
112 U.S. Census Bureau, Statistical Abstract of the United States, 2008 (Washington, DC,
2007), pp. 7, 373; U.S. Department of Transportation, Research and Innovative Technology
Administration, National Transportation Statistics 2007 (Washington, DC, 2007), table 1-
11; U.S. Bureau of Economic Analysis, “Gross Domestic Product,” at
[ ].
113 Pisarski, Alan E., Commuting in America III (Washington, DC, Transportation Research
Board, 2006). Of the 99.1 million commutes originating in a metropolitan area in 2000, 44.3
million (45%) were from suburb to suburb, 25.2 million central city to central city (25%),

18.8 million from suburb to central city (19%), 8.6 million from central city to suburb (9%),

and 2.1 million to a non-metropolitan destination (2%).

64% in 1980. Over the same period, the share of commuters using transit hovered
around 5%.114
These trends have been bolstered by an increase in the number and widespread
availability of motor vehicles. The number of personal motor vehicles (cars, sport-
utility vehicles, pickups, and minivans) per licensed driver passed 1.0 some years ago
and continues to climb. In 2005, the average number of personal motor vehicles per
driver was 1.16. That same year, only about 8% of households were without a
vehicle.115 The low price of gasoline has also contributed to enhancing the
attractiveness of motor vehicles as a transportation option. For about 20 years
beginning in the mid-1980s, the pump price of gasoline was below $2.00 per gallon
(in 2006 dollars) in real terms, lower than at any time from 1918 on.116
Many of these same factors — population and income growth — together with
economic complexity and globalization have led to more demand for commercial
truck transportation. Since 1980, truck traffic has grown slightly faster than
passenger traffic.117 Although a lot of truck milage is made on long intercity trips,
about half of truck VMT is made in urban areas, contributing significantly to urban
traffic congestion, particularly near urban-based industrial facilities, ports, and border
Many of the same factors generating vehicle travel and congestion are expected
to continue growing. The Census Bureau expects the population to reach 364 million
by 2030, an increase of about 20% from 2007.119 Two-thirds of this population
growth, and with it a significant portion of new road traffic, is expected to occur in
just seven states: Florida, California, Texas, Arizona, North Carolina, Georgia, and
Virginia. Over the same period, the CBO projects that GDP will increase by about
70% (in real terms).120 FHWA’s Highway Performance Monitoring System includes
state-based estimates of future VMT growth.121 The annual growth rate is projected
to be 1.92%, with rural VMT growing somewhat faster than urban areas (2.15%

114 Pisarski, 2006; U.S. Census Bureau, 2005 American Community Survey, at
[ h t t p : / / www.c e n s u s . go v/ ] .
115 U.S. Census Bureau and U.S. Department of Housing and Urban Development, American
Housing Survey for the United States: 2005 (Washington, DC, 2006), table 2-7, at
[ h t t p : / / www.census.go v/ pr od/ 2006pubs/ h150-05.pdf ] .
116 American Petroleum Institute, “U.S. Pump Price Update — April 10, 2007,” at
[ h t t p : / / www.api .or g/ about oi l gas/ gasol i n e/ upl oad/ Pump Pr i c eUpdat e .pdf ] .
117 FHWA, 2007, p. 20.
118 U.S. Department of Transportation, Federal Highway Administration, Highway Statistics

2006 (Washington, DC, 2007b).

119 U.S. Census Bureau, 2007, p. 8.
120 Congressional Budget Office, December 2007.
121 Federal Highway Administration and Federal Transit Administration, 2007.

average annual versus 1.79%).122 The Freight Analysis Framework projects that
freight tonnage by truck will double between 2002 and 2035.123
None of this is inevitable, and a few counter trends may slow the growth in
VMT and peak-period travel. For example, although the age at which people are
retiring from the workforce has begun to tick upwards over the past few years, baby
boomers will begin retiring in large numbers in a few years. This may slow the
growth in the number of workers. Some have suggested that as baby boomers age,
they may begin to favor denser neighborhoods that are easier to serve with transit,
thereby reducing the growth in VMT. Others believe there may be a reduction in
work travel associated with flexible schedules, such as a compressed work week and
Interurban Road Traffic Congestion. Most, though not all, road traffic
congestion is experienced in urban areas. An FHWA study of truck travel in freight-
significant corridors — Interstate routes that span urban and rural areas — showed
that a good deal of delay and reliability problems derive from the urban portion of124
trips. Nevertheless, rural travel has grown faster than urban travel during the past

25 years. Between 1980 and 2005, rural VMT per lane mile grew by 65%, whereas125

urban VMT per lane mile grew 41%. Estimates by FHWA of peak-period
congestion on the federally adopted National Highway System in 2002 and a
projection to 2035 suggest a much more widespread congestion problem. In 2002,
FHWA’s analysis of congestion found that it was largely confined to highway links
in large urban areas. However, by 2035, assuming no change in physical road
capacity or operational improvement, FHWA expects congestion to intensify in those126
areas and to spread to intercity corridors throughout the country.
Road Bottlenecks. A number of studies have attempted to locate,
characterize, and quantify bottlenecks in the highway system. TTI defines
bottlenecks as “locations where the physical capacity is restricted, with flows from
upstream sections (with higher capacities) being funneled into them.”127 One study
found 233 major highway bottlenecks in 2002, defined as places with 700,000 hours

122 Ibid., pp. 9-10. Rural VMT is projected to grow faster than urban VMT for several
reasons: urban areas, unlike rural areas, are expected to moderate their VMT growth using
travel demand management techniques; commercial truck travel in rural areas is expected
to grow more quickly than in urban areas; and rural areas include rapidly growing places on
the urban fringe that may be reclassified as urban in the future.
123 Federal Highway Administration, 2007, p. 11.
124 U.S. Department of Transportation. Federal Highway Administration, Office of Freight
Management and Operation, 2006.
125 CRS calculations based on U.S. Department of Transportation, Federal Highway
Administration, Highway Statistics (Washington, DC, annual issues).
126 See the maps in Federal Highway Administration, 2007, pp. 31-32.
127 Cambridge Systematics and Texas Transportation Institute, “Traffic Congestion and
Reliability: Linking Solutions to Problems,” report prepared for U.S. Department of
Transportation, Federal Highway Administration (July 19, 2004), p. 2-1, at
[ eport_04/congestion_report.pdf].

of delay annually. This was a 40% increase in major bottlenecks from the 167
bottlenecks found in 1999. Of the 233 major bottlenecks in 2004, 24 had more than
10 million hours of delay in a year.128 Freeway to freeway interchanges account for
most bottleneck delay. According to another study, highway bottlenecks affecting
large volumes of trucks accounted for 243 million hours of truck delay in 2004.129 A
third study on bottlenecks associated with summer vacation travel ranked the top 25
destinations likely to suffer the worst traffic delay in 2005.130
Road Congestion at International Gateways. Other potential
bottlenecks in the transportation system are foreign trade gateways. Rapid growth
in international trade over the past few decades has placed enormous pressure on
these gateways — land border crossings, certain airports, and water ports — and the
road and rail infrastructure that supports them. By value, in inflation-adjusted terms,131
international merchandise trade increased by 160% between 1980 and 2005.
Growth in value terms has been particularly rapid on the Mexican and Canadian
borders and on the Pacific Coast, although the Atlantic Coast continues to handle the
most trade (Figure 6). These trends are likely to continue with the growing
globalization of production and consumption. Indeed, the FHWA expects foreign
trade tonnage to more than double between 2002 and 2035.132
Although no comprehensive time-series data for congestion at land gateways
nationwide exist, numerous studies have found delay and unreliable travel times at
certain heavily used crossings. In 2004, daytime (8:00 a.m. to 6:00 p.m.) wait times
for trucks entering the United States from Canada averaged 8.5 minutes, and those
from Mexico averaged 7.3 minutes. However, daytime wait times at Laredo, TX,
averaged nearly 21 minutes, and at Port Huron, MI, the average was 25 minutes.133
Although they provide a basis of comparison, these averages mask the variability of
delays that are probably more important. At land border crossings, congestion is
caused by three main problems: inadequate transportation infrastructure to handle the
volume of cars and trucks, import and security processing, and general urban road
traffic congestion.134 Some studies have suggested that border delay and reliability

128 American Highway Users Alliance, 2004.
129 U.S. Department of Transportation, Federal Highway Administration, “An Initial
Assessment of Freight Bottlenecks on Highways,” white paper prepared by Cambridge
Systematics, October 2005, at [].
130 American Highway Users Alliance, American Automobile Association and TRIP, “Are
We There Yet?” (Washington, DC, 2005), at [
131 Federal Highway Administration, 2007, p. 14.
132 Ibid., p. 11.
133 U.S. Department of Transportation, Research and Innovative Technology Administration,
Bureau of Transportation Statistics, Transportation Statistics Annual Report 2005
(Washington, DC, 2005), at [
134 Texas Transportation Institute and Battelle Memorial Institute, “International Border
Crossing Truck Travel Time for 2001,” report prepared for U.S. Department of

problems have more to with institutional and staff issues, such as inspection staffing
levels at periods of high demand, than infrastructure problems, although this may
depend on the specific crossing.135 Similarly, delays at water ports may be caused by
inadequate road and rail infrastructure, general road congestion, and customs and
security requirements. Indeed, one of the big challenges at international gateways in
the past few years has been balancing passenger and freight mobility with the need
for heightened security in the wake of the terrorist attacks of 2001.136
Figure 6. U.S. Merchandise Trade by Region, 1980-2005

Source: U.S. Department of Transportation, Federal Highway Administration, Freight Facts and
Figures 2007 (Washington, DC, 2007).
Measures and Trends of Congestion in Public Transit
The main public transit modes in the United States — bus, commuter rail, heavy
rail, and light rail — have different but overlapping characteristics that influence the
causes and impacts of congestion. All public transit modes have the potential for
vehicle overcrowding, but they differ in terms of system congestion. Transit buses
typically run on roads in the general traffic stream and, therefore, are affected by road
134 (...continued)
Transportation, Federal Highway Administration (April 2002), at [
135 Taylor, John C., Douglas R. Robideaux, and George C. Jackson, “U.S.-Canada
Transportation and Logistics: Border Impacts and Costs, Causes, and Possible Solution,”
Transportation Journal, vol. 43, no. 4, pp. 5-21.
136 Testimony of Margaret Wrightson, Director of Homeland Security and Justice Issues,
Government Accountability Office, in U.S. Congress, Senate Committee on Commerce,
Science and Transportation, May 17, 2005, at [].

traffic congestion. In many cities, light rail systems have their own rights of way, but
running at grade with limited separation can cause conflicts between rail and road
traffic. Commuter rail service runs over rail lines that also carry freight and intercity
passenger trains and, therefore, is subject to many of the same causes of delay and
unreliability. Heavy rail (subway) systems have their own rights of way and, thus,
are not subject to conflicts with other modes. However, subway system congestion
is theoretically possible at peak periods when the number of trains running on the
track begins to reach the design maximum, known as line capacity, and passenger
loads affect station dwell times.137 When running at full capacity, the lack of
redundancy in the system also magnifies the effect of incidents such as a train
Transit ridership grew 15% between 1980 and 2005. Over that time, bus
ridership was virtually unchanged, while commuter rail and heavy rail grew by 51%
and 33%, respectively. Light rail ridership almost tripled during these years because
of the construction of several new systems.138 Although all urban areas and many
rural areas provide some sort of transit service, transit usage is heavily concentrated
in a few large urban areas. Bus transit is widely provided, but only 34 metropolitan
areas have one or more major forms of rail transit (defined here as commuter rail,
heavy rail, and light rail). In 2004, 10 metropolitan areas accounted for 75% of all
urban transit trips in the United States (see Table 2). The New York metropolitan
area alone accounted for nearly 40% of all urban transit trips.
There are no direct measures of public transportation congestion available
regularly on a national basis. Two indirect measures of congestion are average
vehicle utilization, as a measure of vehicle overcrowding, and average operating
speeds, as a measure of system congestion.139 Vehicle utilization, as measured by
the USDOT, is “calculated as the ratio of the total number of passenger miles
traveled annually on each mode to total number of vehicles operated in maximum
scheduled service in each mode, adjusted for the passenger-carrying capacity of the
mode in relation to the average capacity of the Nation’s motorbus fleet.”140 The
USDOT notes that these two variables are related as “changes in the capacity
utilization of rail vehicles have influenced these vehicles’ operating speeds through
changes in dwell times. As vehicles become more crowded, they take longer to
unload and load, increasing wait at stations and hence passengers’ total travel141

137 Transportation Research Board, Transit Capacity and Quality of Service Manual, 2nd
Edition, TCRP Report 100 (Washington, DC, 2003), at [
138 American Public Transportation Association, “Unlinked Passenger Trips by Mode, 1890-

2005,” at [].

139 Federal Highway Administration and Federal Transit Administration, 2007.
140 Ibid., pp. 4-3
141 Ibid., pp. 4-22.

Table 2. Top 10 Metropolitan Areas by Transit Usage, 2004
AnnualaCumulative %
Passenger TripsTransitU.S.
Urbanized areaRank(thousands)TripsPop.
New York, NY-NJ-CT13,383,886386
Los Angeles, CA2606,8434511
Chicago, IL-IN3582,7865214
Washington, DC-VA-MD4442,9365716
Boston, MA-NH-RI5396,0876117
Atlanta, GA6363,3266519
Philadelphia, PA-NJ-DE-MD7350,5186921
San Francisco-Oakland, CA8199,3697122
Seattle, WA9156,2567323
Miami, FL10151,2227525
United States, urban total8,852,131
Sources: U.S. Department of Transportation, Research and Innovative Technology
Administration, Bureau of Transportation Statistics, State Transportation Statistics 2006
(Washington, DC, 2007), table 4-3; U.S. Census Bureau, Statistical Abstract of the United
States, 2007 (Washington, DC, 2007), tables 17 and 25.
a. Unlinked passenger trips is the number of passengers boarding transit vehicles. A transit
trip from origin to destination may involve one or more than one unlinked trips.
Average vehicle utilization data for urban transit systems show that passenger
volumes in relation to service capacity are greatest on rail, particularly commuter rail.
The higher level of commuter rail utilization is due to the longer average trip lengths
with seating capacity only and to the limited time service is available. According to
the FTA, utilization rates have generally declined since 2000/2001 (Figure 7). These
data are bolstered by data on average speed that show little change in the average
speed of non-rail modes, mainly buses, but a slight decline in speeds for rail transit.
Non-rail speeds averaged 13.7 miles per hour in 1995 and 14.0 mph in 2004, but rail142
speeds declined from 26.6 to 25.0 mph over this period. Nevertheless, anecdotal
evidence points to overcrowding problems on some rail transit systems, such as
Washington’s Metro and Boston’s T. This suggests that these national average
utilization data, which average over time and across place, may not fully capture rail
transit overcrowding and system congestion in certain cities at certain times.

142 Ibid., exhibit 4-15.

Figure 7. Transit Vehicle Utilization, 1995-2004

Source: U.S. Department of Transportation, Federal Administration and Federal Transit
Administration, 2006 Status of the Nation’s Highways, Bridges, and Transit: Conditions and
Performance (Washington, DC, 2007).
Measures and Trends of Congestion in Rail
Freight Rail Congestion Measures. The rail network is made up of a
system of mainlines, spurs, sidings, yards, intermodal terminals, and places where the
lines of different railroad companies come together (known as interchanges).
Complexity is added by the physical characteristics of the thousands of tunnels,
bridges, and overpasses with different clearances, the number and type of highway-
rail grade crossings, and the thousands of miles of track with different load-bearing
capacity and parallel lines. For the most part, this railroad infrastructure is owned
and operated by private companies engaged in the transportation of freight.
However, in some places, freight trains share space with passenger trains belonging
to Amtrak and, in some urban areas, commuter rail operators.
In contrast to the way highway transportation works, decisions about
accessing the rail system are controlled by a central authority — each railroad — that
determines when a shipment will be transported and for what price. Thus, capacity
problems tend to appear in a different form than they do on the highways and must
be measured in different ways. Moreover, because the rail system is primarily
private, the government has chosen not to collect and publicly disclose detailed data
related to congestion. As a result, some indications of congestion problems are
impressionistic and anecdotal.

In a free-market, when demand outstrips supply for a good or service, the
price rises until an equilibrium between the two is found. One indicator of
congestion in the rail industry, therefore, is freight rates. Unfortunately,
understanding the relationship between capacity and prices is difficult as best. Rates
are affected by any number of other variables, including the competition of other
modes. Morever, rates can be regulated after the fact to protect “captive shippers.”
Capacity problems may also result in deterioration in service quality or no service at
all. For example, in some cases, there may be a promise to transport a shipment at
a certain price, but this shipment may be delayed as the operating railroad waits for
space on the network. In other cases, some shipments may be denied access to the
system completely and will have to travel by another means of transportation.
In theory, centrally controlled access to the rail system should avoid the
queuing seen on highways; however, in practice, delay and unreliability do tend to
increase as the number of trains on the system reaches maximum capacity. This
derives from the complexity of determining the timing and routing of trains with
different dimensions, such as single- or double-stacked containers, carrying different
commodities over long distances, and the rules that must be followed to ensure that
trains do not collide, particularly in places that are not signal-controlled. In addition,
tight schedules can be upset by unforseen incidents such as accidents, bad weather,
and breakdowns and by interference with passenger trains that, by federal law, are
supposed to have priority over freight trains.
Publicly available measures of freight rail congestion are traffic density,
speed, and freight rates. None of these conclusively proves that congestion is a
problem because they are all influenced by other things, such as efficiency gains
derived from improved technology. Traffic density, as the Association of American
Railroads (AAR) notes, “measures the average system-wide freight carrying
utilization of the railroad track infrastructure. A higher figure indicates greater
utilization efficiency, but can signal the risk of congestion.”143 Speed can be
measured by average train speed or by net ton-miles per train hour (freight speed).
Again, slower speeds might be an indication of a congestion problem, but they might
also be related to other factors, such as the mix of commodities being transported and
length of haul. Average cost is measured by freight revenue per ton-mile. TRB notes
that this has been declining for years because of productivity growth, excess capacity,
and deregulation. It notes a slowing of the rate of decline or even a pronounced
increase might be indicative of a congestion problem.144
Trends in Freight Rail Congestion. The three measures of capacity
utilization — traffic density, average freight speed, and freight rates — all suggest
a growing congestion problem in the industry. This is supported by anecdotal
evidence of trip times and bottlenecks. Since rail deregulation in 1980, Class I rail
freight ton-miles have increased 93%, from 919 billion to 1,772 billion, while miles
of track have decreased 40%. Traffic density measured by millions of revenue ton-

143 Association of American Railroads, Railroad Facts 2007 (Washington, DC, November

2007), p. 42.

144 Transportation Research Board, Freight Capacity for the 21st Century, Special Report

271 (Washington, DC, 2003), p. 62.

miles per mile of track, therefore, has increased from 3.4 in 1980 to 10.9 in 2006
(Figure 8).145 Moreover, these data exclude demands placed on the system by
intercity and commuter passenger rail operations.
Figure 8. Freight Rail Traffic Density, 1980-2006

f T
le o
les Per Mi

19 80 19 85 1 99 0 19 95 2 00 0 20 05
Source: Association of American Railroads, Railroad Facts (Washington, DC, various issues).
The average speed of freight moved by rail, measured by net ton-miles per
train hour, grew substantially in the 1980s but has since declined (Figure 9).
Consequently, as CBO notes, the average speed is “now lower than it has been since
the early 1980s, except for the turbulent 1997-1998 period following the merger of
Union Pacific and Southern Pacific.”146 Another expert estimates that over the past
10 years, trip times have increased by about 25%-50% for general merchandise rail
traffic. 147
145 Association of American Railroads, 2007.
146 CBO, January 2006, p. 8.
147 Testimony of Carl D. Martland, April 26, 2006.

Figure 9. Average Speed of Freight by Rail, 1980-2006

50,000 Tr
Ton-Miles per
1980 1985 1990 1995 2000 2005
Source: Association of American Railroads, Railroad Facts (Washington, DC, various issues).
Average freight rates, measured by freight revenue per ton-mile, have
declined substantially since deregulation from 5.3 cents per revenue ton-mile to 2.4
cents (in constant 2000 dollars). However, over the past decade the decline in rates
slowed, and in the past few years rates have increased. Rates in 2006 were 14%
higher in real terms than they were in 2003 (see Figure 10).148 It is not clear,
however, if this is indicative of a new upward trend in rates, nor is it clear how this
relates to capacity problems in the industry.
Like road traffic congestion, freight rail congestion is generally limited to a
few key locations. Research completed for the Association of American Railroads
indicates that about 3% of the freight rail network has demand at or above capacity,
with another 9% near capacity. Some major bottlenecks include, among others, the
network in and around Chicago, Kansas City, Atlanta, and Memphis as well as the
rail corridors from San Francisco to Los Angeles and Los Angeles to Tucson,
Arizona. In the Chicago region, congestion is compounded by the lack of
connectivity between the several different railroads serving the area whose route
systems are focused on states east and west of the Mississippi River.149
148 CRS calculations using the implicit price deflator for GDP.
149 Association of American Railroads, National Rail Freight Infrastructure Capacity and
Investment Study, Washington, DC, September 2007, at
[ ht t p: / / www.aar .or g/ PubCommon/ Docume nt s/ nat l _f r e i ght _capaci t y_st udy.pdf ]

Figure 10. Average Freight Rates, 1980-2006
(constant 2000 cents)

6. 0
5. 5
5. 0)
4.5000 cents
t 2
4. 0n
3. 5
3.0r Ton-Mile (consta
2.5nue P
2. 0Reve
1. 5
1. 0
1980 1985 1990 1995 2000 2005
Sources: Association of American Railroads, Railroad Facts (Washington, DC, various issues); U.S.
Bureau of Economic Analysis, “Gross Domestic Product.”
Intercity Passenger Rail (Amtrak) Congestion Measures.
Congestion problems in intercity passenger train travel — trains operated by the
National Railroad Passenger Corporation, known as Amtrak — are somewhat akin
to those of the freight railroads discussed above. Except for the 500 miles it owns
in the Northeast Corridor (NEC), intercity passenger trains operated by Amtrak run
on rail lines that are owned and operated by freight railroads. As freight movements
have grown, so too have the conflicts between freight and passenger trains, even
though under existing federal law, passenger trains are supposed to have priority over
freight trains. Other issues for Amtrak include the condition of the privately owned
rail lines that can result in a local speed restriction below the track’s normal speed,
train breakdowns, and other incidents. Measures of these types of congestion
problems are train on-time performance, amount of delay, and average speed.
In addition, as a type of passenger service, congestion problems with Amtrak
theoretically may be manifest in ticket availability, ticket prices, and train
overcrowding. Systemwide, these are generally not issues that Amtrak has to worry
about. These problems may occur on certain routes at certain times, such as the NEC
around major holidays, but realistically, the system cannot be designed to handle
demand that only occurs a few times a year. Load factor, a metric tracked by Amtrak,
is a measure of train utilization and possible overcrowding.
Intercity Passenger Rail (Amtrak) Congestion Trends. The data
appear to show that, in general, rail system congestion, including freight, commuter,
and Amtrak operations, is something of a problem and is getting worse, but that train

overcrowding is not a problem. Amtrak delays per 10,000 miles have trended
upward from FY2001 through FY2006. Delays resulting from Amtrak itself have
remained relatively constant during that period, at about 400 minutes per 10,000 train
miles. Most of the delays are due to freight operations, rising from about 1,700
minutes in FY2001 to about 2,300 minutes in FY2006. Overall on-time performance
was 67.8% in FY2006, down from 69.8% in FY2005, 70.7% in FY2004, and 74.1%
in FY2003. Load factors, on the other hand, are quite low, suggesting little train
overcrowding. For all Amtrak routes, the load factor in FY2006 was 48%. The
average load factor in FY2006 was 45% in the NEC, 41% in state-supported and
other corridors, and 55% on long distance routes.150
The Costs of Transportation Congestion
The negative effects of transportation congestion are primarily economic.
Transportation congestion, particularly road traffic congestion, also causes a good
deal of stress in some of those that experience it, as well as a certain amount of
environmental damage because of the extra fuel that is used. Congestion may also
have a negative effect on road traffic safety, although it is not clear from the available
evidence if the damage done as a result of slowing or stopped vehicles outweighs the
reduction in crash severity due to lower speeds. However, the main effects are an
increase in direct user costs, particularly the extra time and fuel expended, and a
number of economic distortions that decrease productivity and hurt competitiveness.
Most of the available evidence on the costs associated with transportation
congestion is limited to the effects of road traffic congestion. Little is known about
the national costs associated with rail, transit, and intermodal congestion. Hence, if
accurate, existing estimates focusing exclusively on the costs of road traffic
congestion understate the total cost of transportation congestion to the national
economy. It must also be borne in mind that estimates of the cost of congestion are
based on assumptions that are somewhat arbitrary. Time, an important variable in
transportation evaluation studies, can be especially hard to value.151
The direct user costs of road traffic congestion are the extra time and fuel
expended to complete a trip. In its study of 437 cities, TTI estimates that drivers lost

4.2 billion hours to road traffic congestion and wasted an extra 2.9 billion gallons of152

fuel, at a cost of $78.2 billion. Most of the cost is due to the time lost by travelers.
Per traveler, the cost is $710 annually or approximately $3 per work day. In
inflation-adjusted terms, the cost of congestion has risen from $14.9 billion in 1982

150 Amtrak, “Monthly Performance Report for September 2006,” December 4, 2006, at
[]; Amtrak, “Monthly Performance Report for
September 2004,” November 1, 2004, at [].
151 U.S. Department of Transportation, “Departmental Guidance for the Valuation of Travel
Time in Economic Analysis,” memorandum, April 9, 1997; and U.S. Department of
Transportation, “Revised Departmental Guidance,” memorandum, February 11, 2003, at
[ programs a.htm#V].
152 TTI assumes a cost of $14.60 per hour of person travel and $77.10 per hour of truck time.
Excess fuel is estimated using the state average cost.

(in constant 2005 dollars). These estimates, however, do not include the cost of
unreliability, in that travelers will often budget extra time to make sure they arrive
on time, even if it means arriving early.
In addition to direct user costs, there are at least three other types of economic
costs associated with congestion153:
!Logistics costs — the extra costs associated with businesses having
to carry extra inventory as a result of slower and more unreliable
!Market scale and accessibility costs — as congestion reduces the
area that can be served by a production facility, the reduced demand
results in higher unit costs because of lower-scale efficiencies and
lower access to specialized inputs.
!Business cost of worker commuting — the costs associated with
attracting and retaining workers and compensating them for higher
commuting costs. There may also be lower labor productivity
resulting from the stress of longer or more unreliable commutes.
Although not quantified, congestion in other modes also has costs. As
demand for space on the rail system increases, rates may begin to rise, increasing
shipper costs. In addition, railroads have been keen to accommodate generally more
lucrative intermodal shipments over bulk shipments. This is beginning to create
significant problems for the movement of bulk shippers in some markets at certain
times, as they often have no alternative to moving their goods by rail. Congestion on
the rail system may also force more freight to move by truck. Some contend that
there are a number of public benefits associated with moving freight by rail, such as
less air pollution per ton-mile of freight than trucking.154 Similarly, congestion and
overcrowding in passenger rail transportation and public transportation may divert
travelers to other modes. In urban areas, congested transit service may lead to more
single-occupant driving during the peak period, causing more road congestion.
Likewise, congested intercity rail transportation might shift a few travelers onto the
roads, although it may shift them to intercity buses or airplanes, depending on the
It is commonplace these days to attempt to quantify the costs of congestion
and add them together to arrive at a total cost of congestion to the economy,
sometimes expressed as a share of GDP. This approach is particularly common in
accounting for the costs of road traffic congestion, as TTI does in terms of extra time
and fuel, and other researchers have attempted to calculate more comprehensively.155

153 Transportation Research Board, Economic Implications of Congestion, National
Cooperative Highway Research Program, Report 463 (Washington, DC, 2001), at
[ onlinepubs/nchrp/nchrp_rpt_463-a.pdf].
154 AASHTO, 2003.
155 For example, the Chief Economist of the U.S. Department of Transportation adds TTI’s

There are, however, some problems with this approach. These cost estimates are
often based on the premise of “free-flowing traffic,” which, as discussed above, tends
to exaggerate the amount of congestion experienced. Furthermore, total cost
estimates suggest that there is a monetary windfall waiting to be distributed to every
household, when in reality, eliminating congestion, if it were possible, would only
save most travelers a few minutes on peak-period trips.156 Consequently, a number
of experts question the calculation of total costs and suggest that
what matters in practical terms is the change in the cost of congestion
brought about by a specific feasible projects or act of policy.... As
economists would say, we need to change our thinking from total costs to157
marginal costs.
Transportation Congestion Remedies
Transportation engineers and planners have devised a large number of
potential remedies for congestion. Although it is beyond the scope of this report to
evaluate all of these, it is worthwhile discussing some of the major remedies as a
basic guide for policy makers. The many different remedies form three basic
strategies for reducing congestion: adding new capacity, operating the existing
capacity more efficiently, and managing demand. This section discusses these
strategies and the institutional issues that affect the implementation of congestion
remedies. This is followed by a discussion of rail congestion remedies and
intermodalism in freight transportation.
Building New Road and Transit Capacity
Building new roads, or expanding existing ones, is one approach to reducing
congestion. Proponents of road building point out that since the completion of the
interstate system, road construction has generally lagged behind the growth in motor
vehicle travel. Moreover, these proponents argue that in some places, lack of
capacity is a major contributor to road congestion. TTI’s analysis of congestion158
found that adding to road capacity slowed the growth in travel delay. New capacity
can range from major new freeways to major bottleneck reduction projects and much
smaller projects, such as widening arterial roads and improving street connectivity.

155 (...continued)
estimate for 85 urban areas contained in the 2005 Urban Mobility Report with the cost of
urban areas not included and other factors to arrive at a total of $168 billion annually. See
Wells, Jack, “The Role of Transportation in the U.S. Economy,” PowerPoint presentation
to the National Surface Transportation Policy and Revenue Study Commission (June 26,
2006), slide 21, at
[http:// www.t r anspor t a t i onf or t o mo r r o w.or g/ pdf s/ commi ssi on_meet i n gs / 0606_meet ing_
washingt on/wells_presentation_0606_meeting.pdf].
156 Downs, 2006.
157 Goodwin, 2004, p. 14.
158 Texas Transportation Institute, 2007.

Few deny that highway travel has grown more than highway capacity during
the past few decades. There is, however, a major disagreement about whether new
road capacity, in the absence of tolling pricing, can solve congestion because of the
problem of induced demand (see earlier discussion). Other concerns about major
expansions of road capacity have to do with the costs in labor and raw materials,
rights-of-way acquisition in heavily developed urban areas, and social and
environmental disruptions. Over the past few years, the cost of raw materials has
increased dramatically, making this a greater concern than just a few years ago. An
added difficulty is the time it takes to plan, design, and build major new facilities.
Consequently, some experts argue that once congestion has developed, it is very hard
for an area to build its way out of the problem because of the time it takes to add new
Some suggest that road congestion is a problem because other viable means
of transportation are not widely available. In this view, new or expanded public
transportation service is seen as a major solution to urban road traffic congestion.
TTI points out that if public transit service disappeared and everyone used private
vehicles, delay in the 437 urban areas it studied would increase by 541 million hours,
about a 13% increase.159 By its estimates, almost all of this extra delay (about 80%)
would occur in very large urban areas (population of 3 million or more). This is
because, as noted above, transit service is heavily concentrated in just a few major
metropolitan areas. Currently, about 5% of workers commute by transit and in only
the New York and Chicago metropolitan areas do more than 10% of commuters use
transit. Nevertheless, much higher proportions of transit users are found for certain
types of commute, particularly those from suburb to central city. It is probably in
these sorts of situations — where the density of origins and destinations is high
enough to make transit an attractive mode of travel — in which new or expanded
transit options are likely to contribute to a reduction in road traffic congestion.
Morever, because buses can be caught up in road traffic congestion, only dedicated
bus lanes or non-highway modes of transit provide effective solutions. Generally
speaking, transit is not likely to reduce congestion in smaller urban areas or in the
suburbs of large urban areas because the areas to be covered are too large and the
densities of residences and jobs too low.
According to some experts, new or expanded transit systems have improved
travel options but have not noticeably reduced road traffic congestion.160 To some
extent, this is because most new major transit systems are built in fast-growing
regions in which the growth in travel demand tends to swamp the extra capacity.
However, some contend that peak-period road traffic congestion is not reduced
because if some people switch from road to rail others are induced to travel by car at
the most convenient times, or because many rail riders are not former drivers but
former bus riders. Morever, even though, theoretically, with more transit service, a
greater number of people are able to travel at the most convenient times, the new
capacity may not serve the greatest needs, such as suburb-to-suburb commutes.

159 Ibid.
160 Downs, 2006.

Like new highway capacity, new transit capacity is costly in terms of labor,
materials, and, in some cases, right-of-way acquisition. However, transit can have
positive social and environmental benefits, such as potentially greater mobility for
the poor and non-drivers, as well as lower air pollutant emissions per trip. New rail
systems are the most costly, although light rail can be a cheaper alternative than
heavy rail. The cost of new commuter rail capacity depends largely on whether or
not the existing freight rail network is available for use by passenger trains. Because
of the large start-up costs, some proponents of expanded transit capacity argue that
new forms of bus transit, such as bus rapid transit (BRT), are a more viable
Operating Existing Capacity More Effectively
Operational improvements on highways and transit have become a much
more important concern of state and local DOTs as congestion has increased.
Operations include a host of strategies for improving the flow of road traffic and
improving transit trips. These include, among others, transportation management
center operations, incident management techniques, event management techniques,
ramp metering, real-time traveler information, road weather information systems,
work zone management, signal retiming, and transit priority at signals. Many of
these strategies rely on the deployment of Intelligent Transportation Systems (ITS)
In general, operational strategies for reducing congestion can be quicker to
implement and relatively low-cost. For instance, with a large share of road traffic
congestion caused by incidents and other non-recurring forms of delay, many areas
have created transportation management centers to improve the response of state and
local agencies to problems that can arise at any time or place in the transportation
system. Evaluations have shown that in many cases, the benefits of these centers
greatly outweigh the costs.161 Another advantage of these types of programs is that
they typically cause minimal disruptions, unlike major construction projects. On the
downside, operational strategies require a much greater ongoing commitment from
local and state DOTs. This has been a problem in some places because, historically,
DOTs have functioned as road construction and maintenance agencies and have
struggled to redefine their mission.
Managing Demand
Operational strategies reduce congestion on the supply side of the
transportation equation. There are a range of strategies that exist on the demand side,
known as demand management strategies. Among others, these include congestion
(or value) pricing, high-occupancy vehicle (HOV) lanes, alternative work schedule
and telecommuting programs, and land-use strategies. Proponents of demand
management strategies argue that just as adding a few extra cars on a roadway can
make a big difference in terms of extra delay, removing a few cars can make a big

161 U.S. Department of Transportation, Joint Program Office, Intelligent Transportation
Systems Benefits, Costs and Lessons Learned: 2005 Update (Washington, DC, May 2005),
at [].

difference in terms of reducing delay. For example, an evaluation of the congestion
charge in London, described below, suggests that while traffic has been reduced by
about 15%, congestion has been reduced by about 30%.162
Congestion Pricing. Schemes to charge drivers a fee to travel on
congested facilities or in congested areas are known generally as “congestion” or
“value” pricing. Economists generally believe that congestion pricing is the single
most viable way, though not necessarily most popular way, to reduce highway
congestion. With the use of advanced technologies, the fee can be varied to ensure
the most efficient use of the facility.
There are four main forms of road congestion pricing: variably priced lanes,
variable tolls on entire roadways, cordon charges, and variable areawide charge163
pricing. Cordon pricing, like the one instituted in London in 2003, charges a fee
for entering an area at certain times. Facility-based pricing charges a fee to use a
specific facility — usually a freeway or freeway lane — depending on the time of day
and the amount of traffic on the facility. Variable areawide pricing would use some
sort of vehicle tracking technology to charge for the amount of travel and the types
of facilities used over an entire area.
The main advantage of congestion pricing is that demand can be managed to
offer travel that is less likely to be subject to delay, especially unpredictable delay.
Another advantage is that on existing roadways, congestion pricing can be
implemented relatively quickly. Moreover, with congestion pricing, the negative
external effects are minimal and the effects may even be positive, such as a reduction
in air pollutant emissions from idling vehicles. For state and local governments,
congestion pricing provides a revenue stream to pay for building and operating
transportation facilities.
Congestion pricing schemes are often unpopular and have been criticized in
a number of ways. One criticism is that they discriminate against low-income
drivers. Although it is true that the toll will represent a greater burden for drivers
with lower incomes, research has shown that low-income drivers do use tolled
facilities, suggesting that they often value the time saved. Others propose that pricing
facilities ought to be reserved for new capacity, particularly when it is made available
alongside a typically congested but free alternative. Another criticism of congestion
pricing is that by making it more expensive to travel downtown, the types of areas or
facilities most likely to be tolled, businesses and consumers are likely to seek out
locations away from the tolled areas, resulting in more sprawl. Some contend that,
depending on how it is implemented, traffic may be diverted from the newly tolled
facility to other roads that may be less well-equipped to deal with heavy volumes.
Finally, some have argued that charging new tolls on an existing roadway is a form

162 Transport for London, “Central London Congestion Charging: Impacts Monitoring,
Fourth Annual Report,” June 2006, at [
163 U.S. Department of Transportation, Federal Highway Administration, “Congestion
Pricing: A Primer,” December 2006.

of double taxation because users have already financed the construction of the road
through the gas tax and other user fees.
Land Use Strategies. It is often asserted that low density, suburban
growth in housing and employment has contributed to road traffic congestion.
Hence, some have suggested that one approach to congestion is to encourage
different types of land use development that will reduce reliance on single-occupant
vehicle travel. The two main types of land use strategies that are commonly
proposed are (1) to encourage increased housing and employment density and (2) to
improve the jobs/housing balance. The first often comes under the rubric of transit-
oriented development, whereby more density will make transit, walking, and cycling
more attractive transportation options. The second type of strategy does not
necessarily entail alternatives to driving, but driving can be reduced when people live
closer to where they work.
Although these are desirable strategies in many ways, experts point out one
of the main disadvantages of them is that land-use patterns take decades to evolve,
hence decisions taken today will take years to make a difference in the overall
transportation/land-use system. Experience with increasing land-use densities shows
that such strategies are not likely to reduce congestion per se, although they are likely
to increase accessibility. In addition, some have suggested that such policies may
raise the costs of developing housing, offices, and other types of facilities primarily
by making land more expensive. Another disadvantage is that land-use decision
making tends to be highly fragmented, so that policies to slow growth in one
jurisdiction may lead to “leapfrog” development in another jurisdiction, causing more
travel and more congestion. Research on improving the jobs/housing balance shows
that it is unlikely to reduce congestion because, for a number of reasons, it is very164
difficult to get people to live near where they work.
Institutional Issues
The problem of transportation congestion is compounded by the highly
fragmented planning and operation of the transportation system. Most urban areas
comprise numerous local governments, and some span multiple states. Important
interstate corridors, like I-95, by definition suffer jurisdictional fragmentation. Even
in a single jurisdiction, multiple agencies are responsible for different aspects of the
system. Transit systems, for example, are often operated independently of local and
state departments of transportation. Highway incidents may involve a whole host of
agencies, including state police, local police, ambulance, fire, and state and local
DOTs. In some cases, fragmentation involves a public-versus-private dimension.
The rail system is mostly privately owed and operated, which can make it challenging
to institute new passenger rail service, for example. Because of this fragmentation
many anti-congestion strategies require coordination and collaboration functionally,
jurisdictionally, and across the public/private divide.165

164 Downs, 2006.
165 U.S. Department of Transportation, Federal Highway Administration, Regional
Transportation Operations Collaboration and Coordination: A Primer for Working

Efforts to promote voluntary coordination and collaboration between agencies
and jurisdictions are typically uncontroversial. More controversial are solutions that
affect the funding and authority of different jurisdictions in the planning and
programming of transportation improvements. For instance, some have suggested
that because congestion tends to occur on the regional scale, regional authorities,
such as the Georgia Regional Transportation Authority, and metropolitan planning
organizations should be given more power over the planning and operation of the
transportation system vis-a-vis states and localities.166
Expanding Rail Capacity
Building new capacity in freight rail is seen as a way of dealing with
congestion issues, particularly as a host of technological changes that have improved
operational throughput appear to have run their course.167 In reasonable financial
health today, freight railroads are investing to increase capacity.168 However, there
are concerns that this investment is not keeping up with demand.169 A number of
reasons have been proposed for this, many having to do with the uniqueness of
freight railroading as an industry. To begin with, many note that because track and
the accompanying operational systems are so costly, freight railroading is one of the
most capital-intensive industries in America.170 Also, once constructed, railroad track
is fixed in space, representing a huge wager on future patterns of freight movement.
It has been argued that similar risks are borne by the public sector in the trucking, air,
and waterborne freight industries. Furthermore, like most infrastructure
improvements, it takes a relatively long time to respond to market signals that may
change quickly.
Another issue is whether or not railroads can be a solution to road traffic
congestion by taking truck traffic off the highways. Clearly, rail will not be a
solution to roadway congestion if there is insufficient rail capacity. But should public
involvement for building rail capacity be predicated on relieving road congestion?
As it stands today, there are a host of significant barriers to rail relieving road
congestion, including the fact that many industrial facilities are no longer served by
rail spurs, either because they have been built away from them or because the spurs

165 (...continued)
Together to Improve Transportation Safety, Reliability and Security, FHWA-OP-03-008
(Washington, DC, 2003), at [
T E//13686/13686.pdf].
166 Puentes and Bailey, 2005; Downs and Puentes, 2005.
167 Positive Train Control, a way of managing the movement of trains with advanced
information and communications technologies, which has yet to be applied on a large scale,
may have a significant impact on throughput, as well as other operational factors such as
168 GAO, October 2006.
169 Testimony of Federal Railroad Administrator Joseph H. Boardman, in U.S. Congress,
House Subcommittee on Railroads, Committee on Transportation and Infrastructure, The
U.S. Rail Capacity Crunch, April 26, 2006.
170 CBO, 2006.

were taken out during the downsizing of rail capacity.171 However, the main reason
that rail is unlikely to reduce urban road congestion is that in most places, trucks
make up a small part of the traffic stream.172 Nationally, trucks account for 8% of
highway VMT,173 although the effect of a truck on the traffic stream is greater than
a passenger car. On a multilane highway with no grade, a large truck represents 1.7
cars and at intersections between 3 and 4 cars.174
Nevertheless, many support the idea of public funding for expanding rail
capacity because it will improve the speed and efficiency of the freight system by
allowing shipments to bypass urban road congestion. Moreover, many point out
there are a range of public benefits to moving freight by rail, including less wear and
tear on the roads and a possible reduction in air pollutant emissions. A corollary is
that improved rail system capacity may also reduce the conflicts between freight and
passenger trains, improving the speed and efficiency of both systems.
Intermodalism in Freight Transportation
Many of the solutions for intermodal problems in freight transportation
revolve around the connections to truck and rail transportation at water ports. The
issues in these areas are particularly thorny because most ports are located in already
congested urban areas with very limited space for expansion and because, as very
large facilities in the freight system, they have a major impact on their physical and
social environment in terms of pollution and noise, etc. Moreover, ports involve a
complex mix of public and private organizations, blurring lines of responsibility and
public and private benefits.
In this context, a number of improvements have been proposed to increase the
speed with which freight moves through the system at these critical nodes without
unduly affecting nearby residents. These improvements include extended truck gate
hours, congestion pricing of dock facilities, truck appointment systems, expanded
“on-dock” rail connections, truck-only lanes, and the development of inland ports
connected by fast rail shuttles.
Concluding Observations
Like a number of other public policy issues, transportation congestion can be
viewed as a collection of interrelated problems with severe constraints set in a
context of continual change. These types of issues, sometimes called “wicked
problems” in some, mostly non-transportation, public policy circles, often seem
intractable and typically engender a good deal of frustration that nothing is being

171 Bryan et al., 2006, p. 114.
172 Bryan et al., 2006.
173 Federal Highway Administration, 2007.
174 Bryan et al., 2006.

done or that what is being done is ineffectual at best or counterproductive at worst.175
Among other things, these types of problems typically have other characteristics such
as no definitive definition; a wide variety of potential solutions, but intense
disagreements about the preferred ones and about what constitutes success; and,
because of intended and unintended consequences, a situation where each solution
tends to modify the problem in such a way as to make it manifest in a different form
or in a different time or place.
When tackling these types of problems, public policy experts advise that the
traditional linear approach, in which data are gathered, analyzed, and a solution
formulated and implemented, is not workable. By contrast, they suggest that
“solving” wicked problems is usually an ongoing, complex, and chaotic struggle that
often requires incorporating multiple viewpoints and approaches at once. Moreover,
these experts note that when working on solutions, policy makers and planners often
encounter new dimensions of the problem and, therefore, they must be creative and
In terms of transportation congestion, this suggests a few key ideas for policy
makers to keep in mind as solutions to this problem are crafted and pursued in the
future. To begin with, there is no one solution that will ever fully solve
transportation congestion and that, paradoxically, fully solving the problem may be
undesirable because congestion can be a good problem to have in some
circumstances and may also be a choice about how to distribute scarce resources.
Another key idea is that each solution applied to a dimension of transportation
congestion might create other unintended problems along other dimensions that
require new creative solutions. As such, multiple, iterative strategies likely will be
needed, including supply-side and demand-side approaches; approaches that focus
on passenger systems and those that focus on freight, highway strategies, and transit
strategies; and those that promise short-term results and some that promise
improvement in the long-term. Possibly and most importantly, policy makers and
planners should consider that the ultimate goal may not be to reduce or eliminate
transportation congestion per se, but to focus instead on improving passenger and
freight mobility and accessibility.

175 Horst Rittel and Melvin Webber, “Dilemmas in a General Theory of Planning,” Policy
Sciences, vol. 4 (1973), pp. 155-169.
176 Conklin, Jeffery E. and William Weil, “Wicked Problems: Naming the Pain in
Organizations,” Touchstone Consulting Group, undated white paper, at
[ h t t p : / / www.t ouchst t r / wp/ wi ml ] .