C H A P T E R 6

The Transportation Sector:

Energy and Infrastructure Use

E nergy is a $1 trillion industry representing 8 percent of the U.S. economy.

The two biggest consumers of energy from fossil and renewable fuels are
electric power and transportation. While electricity can be generated from
diverse sources—coal, nuclear fission, natural gas, water, petroleum, and
increasingly, wind and sun—98 percent of transportation, whether by plane,
train, ship or automobile, is currently powered by petroleum. The transporta-
tion sector alone accounts for two-thirds of the petroleum consumed in the
United States. Thus, key to understanding the transportation sector is under-
standing the petroleum market, and the ways in which consumers and firms
in the transportation sector respond to changes in world oil prices.
The lack of substitutes for oil means that in the short run, oil consumption
in transportation is particularly unresponsive to price changes. This makes
the economy vulnerable to sudden increases in oil prices. Perhaps more
importantly, the world’s reliance on oil creates an external cost in terms of
national security.
In addition to petroleum, the transportation sector relies on infrastructure.
The United States has close to 4 million miles of roads, bridges, and highways
to support a wide variety of economic and social activity. Over time, however,
demands on this infrastructure have outstripped its capacity. While the miles
of urban roadways built have increased by nearly 60 percent since 1980,
vehicle miles traveled on urban roadways increased by double that amount.
The primary reason for this shortfall is that a well-functioning market that
puts a price on roadway use is largely nonexistent. As a result, traffic in most
metropolitan areas has become increasingly congested, costing both time and
fuel. In 2003 alone, Americans were delayed about 3.7 billion hours and used
2.3 billion extra gallons of fuel (47 hours and 29 gallons per rush-hour
commuter) in stop-and-go traffic. Like the costs exacted by oil use on national
security and the environment, the full costs of congestion are not taken into
account by individuals when they drive: each driver usually decides when and
where to drive based on his or her own private needs and ignores the costs
imposed on others.

125
 This chapter discusses several developments in the use of energy and
infrastructure for transportation, and reviews strategies that have been used to
reduce oil use and better manage the existing infrastructure. Key points in this
chapter are:
• Recent increases in the price of oil and the external costs of oil have led
to renewed interest by markets and governments in the development of
new alternatives. Government can play a role in ensuring that external
costs are taken into account by markets, but ultimately markets are best
suited to decide how to respond.
• Cars and light trucks are the largest users of petroleum. As a result, the
fuel economy of the vehicles purchased and the number of miles that
they are driven have a large effect on oil consumption.
• Congestion is a growing problem in American urban areas. Cities and states
have shown a growing interest in and capacity for setting prices for road
use during peak periods to reduce the full economic costs of congestion.

Fuel Markets and the Transportation Sector

Over the past 15 years, petroleum use in the industrial, utility, and building
sectors has been relatively flat, while petroleum use by the transportation
sector has grown by 27 percent. This trend is expected to continue. While
new, more energy-efficient technology has reduced the energy needs of most
sectors, gains in vehicle engine efficiency have been more than offset by a shift
to heavier, more powerful cars and light trucks, and increases in driving.
Cars and light trucks accounted for 92 percent of U.S. roadway travel in
2006 and account for 62 percent of petroleum devoted to transport.
Department of Energy projections suggest that these modes of transportation
will continue to be important, and that light truck usage will show significant
growth in the years to come (see Chart 6-1). Heavy trucks consume almost
17 percent of the petroleum used for transport. Air, rail, marine, and off-road
vehicles currently account for the remaining 21 percent. Air travel is one of
the fastest growing modes of transportation. Energy consumption for air
travel is projected to increase nearly 46 percent by 2030, or about 620,000
more barrels of oil per day.

126 | Economic Report of the President

Responding to Changes in the Price of Oil
In well-functioning markets, the price of a good or service reflects all of the
associated costs and benefits—for example, the costs incurred in extracting,
transporting, and refining the oil, or the benefits from using gasoline to drive.
The market then uses price to achieve the most efficient level of production
and consumption. Transportation has largely reacted to changes in energy
markets in this way.
High demand for oil, due in part to rapid economic growth in China and
India, has helped push oil prices to record levels. The real average monthly
price of oil to the refiner was $26 between 1986 and 2004 (see Chart 6-2, in
2006 U.S. dollars). In 2004, the price to the refiner began to climb,
approaching $70 per barrel in 2006 (other oil price measures were higher).
For the transportation sector, this is a significant increase in the cost of one of
its primary inputs. Normally, as the price of a good rises, consumers reduce
how much they use. However, it typically takes years before the transporta-
tion sector’s consumption of oil is substantially reduced, in part due to the
lack of easily available substitutes. Eventually, though, consumers do react to
high prices. For instance, hybrid vehicle sales have tripled since 2004, while
light truck sales have fallen by 16 percent.

Chapter 6 | 127
 When high oil prices are sustained, as has been the case recently, the market
shows renewed interest in investing in new technologies for developing alter-
natives to oil and improving vehicle fuel economy. Such research and
development investments tend to recede when oil prices fall. During the
period of high oil prices in the late 1970s and early 1980s, the private sector
invested billions of dollars in energy research and development before the
price of oil declined. A recent study finds that private investment in alterna-
tive fuel technologies again has increased in response to higher oil prices,
doubling between 2004 and 2006, constituting 10 percent of the total invest-
ment in energy. Because of the transportation sector’s delayed response to oil
prices, these increases are likely to continue for some time.
The lack of alternatives to oil also means that sudden major oil supply
changes—such as when oil production in an entire region is unexpectedly
shut down—can lead to large and sudden price increases in the months
following the shock. Since oil trades in a global market, the impact on the
economy from such shocks does not depend on how much we import, only
on how much we consume, and our consumption has been growing. The
market has adapted to this threat by investing in more energy-efficient modes
of production, investing in alternative energy sources, and increasing holdings
of private oil inventories.

128 | Economic Report of the President

External Costs of Oil Use
Prices determine which goods and services are produced in the market-
place. In the absence of government policy (such as taxes or regulations), the
price of a good or service accounts for all private costs incurred by those who
have produced or purchased the product. In the case of oil, this includes
everyone from the oil company that extracts the oil, to the shipper, refiner,
retailer, and driver who fuels her car. In the case of oil, the price reflects most
of the costs, but there are some costs to society that remain unaccounted for.
Eighty-one percent of the world’s remaining proven petroleum reserves are
currently controlled by members of the Organization of Petroleum Exporting
Countries (OPEC) (including Iran and Venezuela) and Russia, and nearly all
of these reserves are controlled by national oil firms. Since oil trades in a world
market, oil consumption anywhere in the world affects the price of oil for
Americans. The importance of oil to the world economy gives the major oil-
producing countries disproportionate diplomatic leverage in world affairs. Oil
resources can also fuel corruption in developing countries. Air pollutants and
carbon dioxide from burning gasoline also contribute to concerns about air
quality, human health, and climate.
The purchase of a gallon of gasoline imposes these national security and
environmental costs on everyone, not just on the buyer and seller. Though
State and Federal gasoline and diesel fuel taxes and regulations help account
for these other costs, many studies suggest that the total external costs of oil
may be higher. Carefully crafted government policy may be a useful way to
account for these additional costs. However, this objective should be balanced
against additional inefficiencies that government involvement introduces into
the market. Once policies are in place that ensure that individuals account for
the full costs of the goods and services they consume—e.g., national security
and environmental concerns—competitive markets are the most efficient
means to determine how goods are produced, as well as which goods are
produced in the future.

Transportation Fuel Supply

Motor gasoline and diesel fuel will continue to be the main sources of
power for cars and trucks in the near future. In 2006, motor gasoline
accounted for 74 percent of fuel used in highway vehicles, and diesel
accounted for 24 percent (alternative fuels made up the remainder). Diesel
cars and light trucks are uncommon in the United States—only 2 percent of
new cars and light trucks sold use diesel engines; the majority of diesel fuel is
used by commercial vehicles.

Chapter 6 | 129
 Ethanol, an alternative fuel, is currently used as an additive in gasoline to
increase octane and help gasoline burn more completely, reducing emissions
of carbon monoxide and other pollutants. In many states and metropolitan
areas, gasoline sold at the pump contains between 2 and 10 percent ethanol,
depending on State requirements. Using such alternatives to oil can reduce
the environmental costs of transportation as well as the national security
consequences of oil use. To further encourage alternative fuel use, a provision
in the Energy Policy Act of 2005 (EPAct 2005) known as the Renewable Fuel
Standard requires a certain quantity of renewable fuel to be used by gasoline
producers each year. In 2006, producers were obligated to use 4 billion
gallons per year; this obligation will gradually increase to 7.5 billion gallons
in 2012 (Americans consumed about 140 billion gallons of motor gasoline in
2006). One of the strengths of this policy is that it does not choose which
renewable fuel to promote, but allows the standard to be met with any renew-
able fuel that accomplishes the goal of reducing oil use. However, it does not
extend to oil alternatives beyond renewable fuels, such as electric cars or
hydrogen fuel cells. The Renewable Fuel Standard also allows imports to
satisfy the standard, allowing U.S. consumers to take advantage of cheaper
production of renewable fuels in other countries, although this is impeded by
an import tariff on such fuels.
A more significant regulatory change has been applied to diesel fuel.
Starting in 2006, diesel fuel sold in the United States is required to have
a sulfur content of no more than 15 parts per million (ppm), down from
500 ppm in the previous standard. This reduction results in the most strin-
gent diesel fuel standard in the world and enables U.S. consumers to purchase
vehicles with engines that meet clean air requirements using clean diesel fuel.
Diesel engines are between 20 and 25 percent more fuel efficient than compa-
rable gasoline engines (even accounting for the fact that a gallon of diesel
contains more energy than a gallon of gasoline). EPAct 2005 also grants tax
credits to buyers of diesel cars that meet stringent emission standards.

Alternative Fuels and Advanced Technologies

To date, changes in petroleum usage have been driven primarily by the
increasing price of oil and by regulatory concerns. The greatest potential for
large reductions in gasoline consumption stems from new technologies that
could transform how transportation is powered. Over 1 million advanced
technology cars and light trucks were sold in the United States in 2006. About

130 | Economic Report of the President

two-fifths of these were flex-fuel vehicles that can use conventional gasoline
or an alternative fuel called E85, which is approximately 85 percent ethanol
and 15 percent gasoline. U.S. consumers also purchased 256,000 hybrid vehi-
cles in 2006. Hybrid vehicles use an electric motor in conjunction with a
gasoline engine to increase fuel economy.
Use of advanced technology vehicles in the United States is projected to grow
over time (see Chart 6-3). The Department of Energy projects that over 3
million advanced technology vehicles will be sold in 2015 and that by 2030
they will make up more than 25 percent of all light-duty vehicles sold. Of these
advanced technology vehicles, 71 percent are expected to be either
gasoline–electric hybrids or vehicles that can be powered by ethanol and other
plant-based fuels. Though alternative fuels currently power only a small fraction
of our transportation needs, private-sector investments combined with govern-
ment policies are expected to fundamentally change the energy landscape.

Chapter 6 | 131
 Ongoing research explores a wide variety of vehicle fuel technologies such
as electricity, hydrogen fuel cells, and biofuels. Significant technological
barriers exist that prevent the development of these as commercially viable
alternatives. For instance, the wide-scale deployment of hydrogen fuel cells—
devices that combine hydrogen with oxygen in the atmosphere to yield
electricity—will depend on reductions in expense and weight as well as on the
development of clean, cost-effective sources of hydrogen.
Private markets tend to underinvest in innovation of all kinds because
inventors only capture a fraction of the benefits from discovery.
Underinvestment is particularly likely for basic scientific research where
the application to the marketplace may not be evident at early stages.
Underinvestment is also likely when the results of research mainly reduce the
external costs of consumption (such as national security and environmental
costs associated with oil) instead of directly benefiting consumers. In
response, the President’s Advanced Energy Initiative proposed an increase in
annual funding for alternative energy research of 22 percent for fiscal year
2007, adding to the $10 billion of government spending devoted to such
research since 2001.
Several studies find that Federal research and development (R&D) invest-
ment in energy has yielded sizeable societal benefits, not only in economic
terms, but also in terms of knowledge creation and pollution reduction. Still,
the government’s ability to predict which technologies will best meet a given
goal is questionable, so the most effective government policies allow the
market to choose the path of innovation.

Demand for Transportation Fuel

The United States is a vehicle-dependent society. More than 9 out of 10
American households own at least one vehicle, and most households own two.
In 2004, vehicles in the United States traveled close to 3 trillion miles, up more
than 20 percent from 1995. Commuting and other business-related activities
account for about 35 percent of vehicle miles traveled (see Chart 6-4).
Americans also use their cars and trucks to go shopping (15 percent of miles
driven), attend to personal and family business such as medical appointments
and dropping children off at school (25 percent of miles driven), and for social
and recreational activities, including vacations (22 percent of miles driven).

132 | Economic Report of the President

 In spite of widespread vehicle use, the proportion of the American
household budget spent on transport fuel is small (less than 4 percent). That
said, Chart 6-4 shows that a significant share of vehicle miles traveled are
related to nonwork activities, indicating that households may have some flex-
ibility to quickly adjust when the costs of travel are high. In response to higher
prices, drivers make two adjustments: they drive less and they purchase more
fuel-efficient vehicles. Several studies have found that these two effects
combined imply that a 10 percent increase in the price of gasoline will result
in about a 4 percent decrease in gasoline consumption in the long run.
Compared to other commodities, households’ gasoline consumption may
take several years to respond to price changes.
State and local initiatives that encourage use of mass transit and carpooling
focus on encouraging people to drive less. In New York City, the most densely
populated of all cities in the United States, mass transit accounts for
45 percent of all commutes into the central city. New York, however, is

Chapter 6 | 133
unique. Many U.S. cities, such as Phoenix and Los Angeles, are spread out
over a large area, making it difficult to design mass transit corridors that
effectively meet the commuting needs of travelers. Public transportation also
has difficulty competing with the flexibility and convenience of car travel in
these types of cities. In the entire United States, 5 percent of commuters rely
on public transportation.
One way many urban areas try to encourage carpooling is through the
designation of high-occupancy vehicle (HOV) lanes. This method rewards
carpooling by allowing vehicles with two or more passengers to travel in lanes
not open to vehicles with only one person in them. In this way, HOV drivers
can reduce travel time when roads are congested. Unfortunately, HOV lanes
are often underutilized and the popularity of carpooling is not increasing. In
2000, 90 percent of American commuters drove to work each day, but of
these drivers only about 13 percent carpooled, down from almost 20 percent
in 1980. This trend makes it unlikely that initiatives focused on carpooling
will make large strides in reducing vehicle fuel use.

Improving Fuel Economy

Evidence shows that drivers switch to more fuel-efficient vehicles in
response to higher gasoline prices. One study finds that higher gasoline prices
accelerate the retirement of older, less fuel-efficient vehicles, and shift new
purchases toward more fuel-efficient vehicles. Government policies have also
been used to influence vehicle fuel economy. The Corporate Average Fuel
Economy (CAFE) standard, passed in 1975, mandates a minimum mile per
gallon (mpg) requirement for each manufacturer’s fleet of new cars and a
minimum requirement for each manufacturer’s fleet of new light trucks. If a
given vehicle is less fuel efficient than the requirement, the manufacturer
must offset it by producing a vehicle that is more fuel efficient, so that the
average fuel economy for all cars (or for all trucks) the manufacturer sells is
above the required miles per gallon level. One rationale used to justify
increasing the stringency of the CAFE standard is to further induce improve-
ments in the fuel economy of vehicles sold to consumers, reducing the
demand for transport fuel and the external costs associated with oil use.
It is important to note that while improvements in fuel economy translate
into gasoline savings, it is not a one-to-one relationship. Higher CAFE stan-
dards encourage increased driving. Since higher fuel economy vehicles can go
the same distance using less gasoline, the cost of driving a mile is reduced. As
the per-mile cost of driving declines, the quantity of miles driven by individ-
uals tends to increase. This “rebound effect” reduces potential fuel savings
from improvements in fuel economy by 10 to 30 percent. Recent estimates
suggest that as incomes grow, driving decisions will depend less on the cost of
driving, and therefore, the rebound effect is expected to shrink in the future.

134 | Economic Report of the President

 In 1978, CAFE mandated 18 mpg for cars and 17.2 mpg for light trucks.
The CAFE standard became increasingly stringent until 1990, after which it
remained virtually unchanged. It only recently became more stringent for
light trucks. Currently, the CAFE standards are 27.5 mpg for cars and
22.2 mpg for light trucks (including SUVs). The Federal government has
increased the CAFE standard for light trucks through two separate regula-
tions, raising it in increments each year beginning in 2005. By 2011, new
light trucks will meet a 24 mpg standard, reflecting a 16-percent increase.
Also by 2011, the largest SUVs—those weighing between 8,500 and 10,000
pounds—will be subject to the CAFE standard for the first time. The
Department of Transportation based the new standard for light trucks on
vehicle footprint, a measure of size, in line with a recommendation by a
National Academy of Sciences panel as a way to mitigate safety concerns. The
footprint-based CAFE standard for light trucks is also an improvement over
its previous configuration because it ensures that all manufacturers make fuel
economy improvements instead of only those producing a wide mix of vehi-
cles. The Department of Transportation is seeking similar authority to
reexamine CAFE for new passenger cars (see Box 6-1).
The fuel economy of new vehicles rapidly increased over the first 8 years of
CAFE. In part, this was a market response to the dramatic increase in gaso-
line prices between 1973 and 1981. By the late 1980s, however, overall fuel
economy had stagnated. While the fuel economy of cars has continued to
slowly increase over time and has been above the CAFE standard since 1986,
consumers have bought an increasing number of SUVs and light trucks whose
fuel economy has remained close to the mandated level of the light truck stan-
dard. Half of all vehicles sold in 2005 were light trucks, including SUVs,
compared to 20 percent when CAFE was first put in place. This shift in
consumer preferences is a rational response to more than a decade of low real
gasoline prices, rising household incomes, and incentives created by CAFE
requirements. Manufacturers also responded to changing consumer prefer-
ences and CAFE requirements. For instance, while station wagons and
minivans have similar fuel economies, the former are counted as cars, and the
latter are counted as light trucks. In the late 1980s, many manufacturers took
advantage of the difference in the stringency of CAFE standards across cars
and light trucks to phase out the station wagon—a relatively fuel-inefficient
car—and replace it with the minivan—a relatively fuel-efficient light truck.
This shift improved the individual fuel economy of both the car and light
truck fleets but did little to change overall fuel economy. While the CAFE
standard has contributed to improved fuel economy since its inception,
understanding its precise impacts and its interaction with gasoline prices is a
matter of some debate. A recent National Academy of Sciences study also
finds that CAFE may have led manufacturers to produce smaller and lighter
cars, posing a risk to safety.

Chapter 6 | 135
 Box 6-1: The President’s New Energy Initiatives

The President has announced several energy initiatives designed to

increase the country’s energy security by reducing projected gasoline
consumption in the light-duty vehicle transportation sector by 20 percent
within a decade.
About three-fourths of this goal will be met by greatly increasing and
expanding the Renewable Fuel Standard. The new standard will
mandate that 15 percent of transportation fuels come from alternative
fuels. In 2006 about 3 percent of fuels used in light-duty vehicles were
not petroleum-based. Under the revised standard 35 billion gallons will
be alternative fuels in 2017. This initiative reflects the belief that techno-
logical change is the key ingredient to diversifying America’s energy
portfolio. Energy security will increase as the dominance of oil use in
the transportation sector diminishes.
The standard will continue to allow refiners, importers and blenders
to use renewable fuels to meet the standard but will expand to allow for
current or future viable alternatives to petroleum to compete.
Expanding the alternatives that meet the standard makes it easier for
blenders and refiners to comply and affords the market broad flexibility
to find the most cost-effective non-petroleum-based fuel options. In the
event that production of alternative fuels proves more costly than
expected, the President has built in two safety valves to protect
consumers. First, the Administrator of the Environmental Protection
Agency, and the Secretaries of the Department of Energy and the
Department of Agriculture will have the authority to waive or modify
the standard if refiners and blenders have difficulty finding alternative
fuels for purchase. Second, an automatic mechanism will be in place to
prevent the price of gasoline from rising above a threshold due to this
policy. These two provisions ensure a degree of market stability as use
of alternative fuels expands in the marketplace.
The 20 percent goal will also be met through increasing the fuel
efficiency of automobiles. This will occur through reforming and
modernizing CAFE standards for cars and further increasing light truck
and SUV standards. These changes are predicted to reduce consump-
tion of gasoline by an estimated 5 percent, based on the assumption
that increases in the standard of 4 percent each year starting in 2010 for
cars and 2012 for light trucks prove warranted. Three reforms are key to
the President’s proposal of increased stringency of CAFE. First, paral-
leling recent changes for light trucks, the law for cars should be
changed to allow the standard to be based on a vehicle attribute (such
as footprint) to address safety concerns. Second, CAFE for both cars
and light trucks should allow manufacturers the option of increased
flexibility in how they meet the standard, by allowing them to trade

136 | Economic Report of the President

 credits. Any manufacturer that increases fuel economy by more than
what is mandated could generate credits that other manufacturers
could purchase to reduce their costs of meeting the standard. The
benefit of trading credits is that it allows the same overall goal of
improved fuel economy to be met at a lower cost. Third, the rate of
increase of the CAFE standard as well as how fuel economy improve-
ments will be divided between cars and light trucks should be at the
discretion of the Secretary of Transportation, as is currently done for
light trucks. The Department of Transportation will employ the regula-
tory process to determine these increases based on sound science and
an assessment that balances the costs and benefits.
The President has also proposed a new $175 million initiative to give
State and local governments the opportunity to explore innovative
ways—such as roadway pricing and increased use of real-time traffic
information—to reduce traffic congestion and save fuel.
In addition to improving the nation’s energy security profile, these
initiatives will also produce significant benefits by reducing air toxics
associated with petroleum-based fuel. They will also help confront the
challenge of climate change by potentially stopping the projected
growth of carbon dioxide emissions from this sector.

Transportation Infrastructure and

Management of Existing Traffic Flow
In addition to its reliance on oil, the transportation sector also relies heavily
on the existing infrastructure of roads and highways. Under the Intermodal
Surface Transportation Efficiency Act of 1991, the Federal government plays
an important role as overseer of the National Highway System to ensure that
the highway system is “economically efficient and environmentally sound,
provides the foundation for the Nation to compete in the global economy,
and will move people and goods in an energy-efficient manner.” In recent
years, however, the road and highway infrastructure has not kept pace with
the number of miles driven in the United States. When more people use a
roadway than the capacity for which it is built, traffic slows. Commercial
trucking—the most common method of moving freight across the United
States—is increasingly reliant on urban interstate highways, many of which
are congested. Between 1982 and 2003 the share of roads in U.S. urban
areas that are congested rose from 34 percent to 59 percent. Changes in

Chapter 6 | 137
commuting patterns have also spread congestion to more roads. The traditional
suburb-to-city commute has diminished in importance: As of 2000, half of all
commuters drove to jobs in the suburbs, while only 20 percent drove to jobs
in central cities.
Congestion is defined as the marked slowing of traffic as a roadway reaches
capacity. Congestion in the United States manifests itself primarily as a bottle-
neck on a roadway (see Chart 6-5). A bottleneck is a hindrance to vehicle
movement because it involves delays at key intersections, backed-up traffic, or
narrow or obstructed sections of a roadway. Unexpected events such as acci-
dents or other traffic incidents also cause congestion on crowded roadways.
Together, they are responsible for 65 percent of all congestion.

It is important to note that roadways are not congested at all hours of the
day. For instance, on one particular roadway in the Seattle area, a trip that
occurs prior to 6 a.m. or after 10 p.m. takes about 10 minutes (see Chart 6-6).
That same trip takes about 30 percent longer at 8 a.m. and almost twice as
long at 6 p.m. due to slowing traffic. This general trend appears in many U.S.
cities and suggests that it is the timing of vehicle miles traveled more than
their growth that is at the root of the congestion problem.

138 | Economic Report of the President

 One underlying reason why congestion exists on U.S. roadways is the lack
of a private market to price roadway use. Most roads in the United States are
provided by the government, are open to all, and are free of charge.
Economists generally believe that a good may be better provided by the
government when it is difficult for private markets to charge for its use.
Because one motorist’s use of a congested road reduces the road’s value for
other drivers and drivers can be selectively prevented from entering the
roadway through the use of gates or technologies that monitor use, it is
increasingly appropriate to charge drivers for some roadway use in the same
way the private market charges for other goods and services.
A driver decides which road to use based on private needs: for instance, the
shortest distance or fastest route between destinations, or the closest, most
accessible highway. The fact that each driver decides on a route independently
of other drivers is not a problem when the number of drivers is well below the
roadway’s capacity. However, when drivers have free access to roads, crowding
occurs at times of high demand, decreasing vehicle speed and flow. Each addi-
tional driver slows down other drivers on the roadway, causing them to lose
time and to burn extra gasoline. However, drivers typically do not consider
the added costs they impose on others. This is a “get in line” or “queuing”
approach to allocating road space. When there is a shortage of something—

Chapter 6 | 139
for instance, space on a ski lift, or attendants at the Department of Motor
Vehicles—those willing to get in line and wait eventually receive what they
want. This approach to road-use management is inefficient because it allo-
cates road space to those with the time to wait in traffic, not necessarily to
those who value its use most highly.
If a roadway is priced—that is, if drivers have to pay a fee to access a partic-
ular road—then congestion can be avoided by adjusting the price up or down
at different times of day to reflect changes in demand for its use. Road space
is allocated to drivers who most highly value a reliable and unimpaired
commute. This arrangement encourages drivers to consider the tradeoff
between the price of using the road and the additional time and inconven-
ience of using a nonpriced, alternate route, or driving at a noncongested time.
Drivers who place a high value on the predictability and reduced time of
commuting, for instance, a doctor who has been called to the hospital for an
emergency, have the option to pay for access to noncongested roads. Drivers
with more time flexibility, for instance a person doing his or her grocery shop-
ping, can avoid the road and the fee. They can use alternative but more
congested roads, shift when they drive to nonpeak hours, or use mass transit
when it provides a cheaper alternative to driving. The average cost to each
driver falls because drivers have a choice in how they pay for roadway use, in
time or in money.

The Cost of Congestion

Over time, slowing traffic exacts heavy costs on drivers. On average,
congestion caused 47 hours of delay for U.S. commuters and commercial
truck drivers in 85 urban areas during peak hours in 2003. For America’s
13 largest cities, this number is much higher: 61 hours. Extra fuel is
consumed on congested roads because of the effect that waiting in stop-and-
go traffic has on fuel economy. In 2003, sitting in traffic wasted about
2.3 billion gallons of fuel, or almost 1.4 percent of all fuel consumed by light-
duty and commercial vehicles that year. Waiting in traffic can also increase the
cumulative amount of pollution emitted from a vehicle’s tailpipe, which
contributes to poor air quality and more greenhouse gas emissions.
Aggregating over the 85 most congested U.S. cities, the cost of time wasted
in traffic and extra fuel consumed by commuters and commercial truck
drivers due to congestion is estimated to have exceeded $63 billion in 2003
(see Table 6-1). In Los Angeles, the city with the worst congestion, the fuel
and time cost of waiting in traffic was calculated to be almost $1,600 per trav-
eler in 2003. In Philadelphia, congestion is noticeably less than in Los
Angeles, but the estimated cost to travelers is still high: $641 per traveler per
year. In addition, businesses that rely on regular and on-time delivery of
supplies have begun to maintain larger inventories to safeguard against

140 | Economic Report of the President

 TABLE 6-1.— Cost of Congestion in Wasted Time and Fuel in the largest Urban Areas

Annual delay per Total cost Cost per peak

Metro area
traveler (in hours) ($ in millions) traveler

Los Angeles–Long Beach–Santa Ana CA .................... 93 $10,686 $1,598

San Francisco–Oakland CA......................................... 72 $2,605 $1,224
Washington DC–VA–MD.............................................. 69 $2,465 $1,169
Atlanta GA................................................................... 67 $1,754 $1,127
Houston TX.................................................................. 63 $2,283 $1,061
Dallas–Fort Worth–Arlington TX................................. 60 $2,545 $1,012
Chicago IL–IN.............................................................. 58 $4,274 $976
Detroit MI.................................................................... 57 $2,019 $955
Miami FL ..................................................................... 51 $2,486 $869
Boston MA–NH–RI....................................................... 51 $1,692 $853
Phoenix AZ .................................................................. 49 $1,294 $831
New York–Newark NY–NJ–CT ..................................... 49 $6,780 $824
Philadelphia PA–NJ–DE–MD ....................................... 38 $1,884 $641

Source: Texas Transportation Institute, 2005 Urban Mobility Report.

unanticipated delays caused by congestion. A recent study conducted by the

Department of Transportation confirms that congestion has resulted in higher
transportation prices and less reliable pickup and delivery times for freight.

Building More Roads

Expanding road capacity may be an important component of any long-
term strategy to accommodate traffic growth in urban areas. However, there
are a number of reasons why a construction-only strategy to alleviate conges-
tion is likely not the best solution. First, increasing capacity can take years to
complete and is expensive—one study found that a lane costs between
$1 million and $8.5 million per mile to build. Second, new lanes are often
needed in densely populated areas, but these are often also the areas where it
is most difficult to find unoccupied space for expansion, making new lanes
politically controversial. Third, a body of evidence suggests that the addition
of a nonpriced lane to an already congested roadway may do little to alleviate
congestion. This happens for two reasons: new roads generate additional
traffic as drivers take trips to destinations that previously took too long to
reach. And since traffic flow improves initially, drivers who were previously
using alternative, often less congested routes now find the highway with the
added lane more attractive. Drivers continue to redistribute themselves across
the various routes until the costs of using the new route and the costs of using
the existing route are about equal. At this point, no driver can be made better
off by changing routes. Ultimately, the reason why building more roads is
insufficient is because it does not address the underlying problem: roads are
not priced and are therefore subject to overuse.

Chapter 6 | 141
Pricing Road Space
There is reason to believe that reductions in traffic congestion would be
relatively easy to attain. Small changes in the number of cars using a partic-
ular roadway at a given time can result in large improvements in the flow of
traffic. For instance, the addition of just a few school buses makes traffic flow
noticeably worse on the first day of school, while traffic flow is noticeably
better on some State holidays when only a small number of residents stay
home from work.
Congestion pricing dampens demand for roads during peak hours and
spreads usage over a longer time period. Differentiating the price of a good by
the time of day effectively allocates limited space during periods of higher
demand. This approach is used by many providers of goods and services:
movie theaters charge more in the evening than they do midday; ski runs
charge more during weekends than they do on weekdays; airlines raise prices
on tickets during peak seasons; taxi cabs charge more during rush hour; and
railroads often charge lower prices for offpeak traveling.
In addition to improved allocation of road space, charging a fee also
provides urban planners with useful information about when and where to
invest in the expansion of existing road capacity. Expansion should be focused
on roads where drivers demonstrate a willingness to pay that is higher than
the costs of construction. Revenues from roadway pricing may also prove a
viable alternative to taxes as a way to fund the building of new roads in urban
areas. As is the case in other markets, those who use the roadway would pay
for its maintenance and expansion.
In general, there are two ways to price road space to address congestion:
cordon pricing and roadway pricing. Cordon pricing charges a toll to vehicles
for access to a congested area regardless of which roads in the area are used. It
is typically in effect during the work week and varies by time of day. Cordon
pricing has been implemented in a number of cities including London,
Stockholm, and Singapore. While cordon pricing has been considered for
several cities in the United States, it has not yet been implemented here. It is
likely to be less effective in cities that are less dense, do not have adequate
public transportation systems, and have multiple areas of centralized
economic activity (such as Phoenix or Los Angeles).
Evidence suggests that cordon pricing fees have been effective in reducing
congestion where they have been tried. After the first year that cordon pricing
was imposed in London, for instance, congestion fell by 30 percent, average
vehicle speed increased by 20 percent, and bus travel became more reliable
(see Box 6-2). One important mechanism for reducing congestion appears to
be the ability to substitute some form of public transportation for driving.

142 | Economic Report of the President

 Box 6-2: Cordon Pricing Experiences in London and Stockholm

In London, drivers pay an 8-pound fee for daily access to a portion of

downtown between the hours of 7:00 a.m. and 6:30 p.m. on weekdays.
There are no toll booths around the perimeter of this area. Instead,
cameras record the license plates of vehicles and check them against a
list of prepaid vehicles. Drivers have a variety of choices in how they
pay: they can pay at designated service stations, through the Internet,
by text message or phone, or by mail. Weekly and monthly charges also
are available for regular commuters. If drivers have not prepaid, they
have until midnight of the next day to do so. Anyone who drives within
the zone without paying during this time period is fined 100 pounds
through an automated system.
Stockholm also recently implemented cordon pricing, but it differs from
the London system in two ways. First, it charges vehicles via a card
mounted on the windshield that is read electronically by roadside beacons
when cars drive past them. Second, Stockholm uses a variable pricing
system, which means that the fee is higher during rush hour periods.
A recent report on the London policy indicates that cordon pricing
has led to a 30 percent reduction in delay time for city commuters.
Initial reports from Stockholm’s 6-month test period indicate that there
were decreases in traffic of about 22 percent due to cordon pricing.
Large reductions in London and Stockholm traffic were due in part to
increased use of bus transit. In spite of early criticism from drivers and
businesses within the central city, cordon pricing has grown in popu-
larity in London. In Stockholm, this has also been the case: a majority
of residents voted to retain cordon pricing after the test period ended.

Roadway pricing aims to limit congestion on certain routes by charging

variable fees (tolls) to access a particular lane or road, regardless of the final
destination. Ideally, road tolls should be responsive to the actual level of conges-
tion at each moment. By increasing the fee during periods of high demand and
reducing it during periods of low demand, the variable tolls reduce congestion
by encouraging offpeak driving and the use of alternative routes.
Variable tolls are rare in the United States. Most of the over 5,000 miles of
toll roads in the United States have flat tolls designed to generate revenue,
rather than variable tolls to relieve congestion. Where they do occur, they are
typically limited to a single road or freeway. On the congested bridges and
tunnels connecting New York and New Jersey, tolls are discounted by

Chapter 6 | 143
20 percent ($1.00) during nonpeak hours. Results of a small survey indicate
that about 7 percent of drivers changed their behavior as a result of these vari-
able tolls. The most common changes were to switch to mass transit, carpool,
or to increase offpeak driving.
Recently, the Department of Transportation helped fund a small pilot
project in Seattle to examine how drivers would respond if the entire road
system in the city were subject to a variable tolling system. Where and when
participants drove was automatically tracked and transmitted by a device
installed in their car. Participants received prepaid accounts between $600
and $3,000 to pay the tolls. At the end of the pilot, they were allowed to keep
whatever they did not spend. Tolls ranged from 5 to 50 cents per mile
and varied by road and time of day. Preliminary results show that nearly
80 percent of participants decreased the amount they drove or changed when
they drove. On average, participants took 5 percent fewer trips by automobile
and drove 2.5 percent fewer miles each weekday due to tolls. Participants took
10 percent fewer trips and drove 4 percent fewer miles during the morning
commute.
Currently, there are about six U.S. highways that use high-occupancy toll
(HOT) lanes, many of which incorporate variable pricing and were piloted
using Federal funds. HOT lanes are variations of the high-occupancy vehicle
(HOV) lanes discussed earlier in the chapter, but they have greater potential to
reduce congestion since they are less likely to be underutilized. Similar to
HOV lanes, they allow carpoolers to use the road for free or at a discount but
charge a toll to single occupancy drivers for access. The toll frequently varies
by time of day. Some tolls set variable prices based on historical highway use
and adjust rates monthly or quarterly. Other tolls use real-time information on
congestion conditions to adjust tolls dynamically over the course of the day. In
locations where HOV lanes are underutilized, conversion to HOT lanes is
suggested as a way to increase use and to provide more choice to drivers. For
instance, in San Diego, conversion of HOV lanes to HOT lanes on a portion
of Interstate 15 increased usage by 64 percent over a 3-year period. Several
studies confirm that there are substantial gains in societal welfare from
allowing solo drivers to pay for access to existing HOV lanes. Others caution,
however, that when only one HOV lane is converted to a variable toll and
other lanes are free of charge, any temporary decrease in congestion on the
remaining free lanes may be offset by the redistribution of traffic.
The use of real-time or historically based variable tolling on HOT lanes may
have a significant effect on traffic flow. For instance, San Diego’s variable toll
uses real-time pricing, which changes every 6 minutes to reflect the amount of
traffic on the road. Computerized electronic signs make information on the
toll amount and the speed and flow of traffic available to drivers before they
have to decide between the free and priced lanes. Results show that travel times

144 | Economic Report of the President

vary little on San Diego’s variable toll lanes because free-flow conditions are
almost always maintained. In Orange County, the tolls vary by hour and day
of the week, but are based on historical information. While they are adjusted
several times each year, the toll does not convey actual conditions to drivers,
only average conditions. Thus, unexpected events such as accidents can cause
major delays on the variable toll lanes and because drivers do not have up-to-
date information on road conditions, travel time is less predictable.
Despite their potential benefits, toll lanes are sometimes portrayed as
“Lexus Lanes.” The contention is that tolled roadways supply faster routes
only to high-income drivers who can afford to pay the tolls, while lower
income drivers continue to be stuck in traffic. One study finds that drivers
with higher incomes tend to use HOT lanes more often than lower income
drivers, but that lower income drivers rely on toll lanes when on-time arrival
at their destination is important. For instance, you can imagine a case where
a parent is running late, but needs to be at the daycare to pick up his or her
child by a certain time. If the parent is late, and the daycare fines him or her
$10, then paying a $4 toll to arrive on time saves $6. A recent survey also
finds that support for or opposition to HOT lanes is unrelated to income.
Another study finds that lower income, bus commuters were some of the
largest beneficiaries of cordon pricing in London. Bus riders are exempt from
paying the cordon fee, but their commute times greatly improved. Not
surprisingly, the number of bus passengers during morning hours increased.
Experts note that implementation of congestion pricing faces less resistance
where motorists are unaccustomed to free and unrestricted roadway access.
For instance, it may be more feasible to implement congestion pricing on a
new road than on an existing road. Likewise, it may be easier to convert HOV
lanes to HOT lanes. The advent of new technologies that electronically
charge the toll by sensing a microchip placed on the windshield of the vehicle
eliminates the need for a driver to stop and physically pay the toll. These are
increasingly used to charge drivers tolls on existing roadways, making conges-
tion pricing systems easier and less costly to implement.
Historically, one of the largest hurdles to variable price tolling on roadways
in the United States has been the Federal-aid highway program, which has
prohibited states from collecting tolls on interstates or other roads that receive
Federal funding. Federally funded pilot projects that explored variable price
tolling brought the advantages of congestion pricing to the attention of poli-
cymakers. Policymakers also began to explore the use of pricing mechanisms
to reduce congestion in other contexts, such as for allocation of runway access
at airports (see Box 6-3). A transportation bill signed into law in 2005 (The
Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for
Users) provides states with increased flexibility to use tolling to manage
congestion and finance infrastructure improvements, and provides ways to

Chapter 6 | 145
participate in pilot demonstrations of variable tolling. States such as Texas and
Colorado have passed laws allowing the formation of toll authorities at local
levels that can then construct and operate toll roads. States such as
Washington, California, Florida, and Minnesota have identified candidate
freeways for variable tolling.

Box 6-3: Airport Pricing to Decrease Congestion

Though traffic jams are easily observable manifestations of conges-

tion, flight delays and runway bottlenecks also waste time and fuel.
Landing fees at most U.S. airports are directly related to the weight of
the plane, even though lighter and heavier planes tend to consume
approximately the same runway time. This contributes to airport
congestion because it encourages smaller, lighter planes (which can
use smaller satellite airports) to overuse the airport, displacing larger,
heavier passenger planes and reducing the number of passengers that
an airport can serve at a time.
A short-lived experiment at Boston’s Logan airport in 1988 demon-
strates how a change in the landing fee structure can effectively reduce
airport congestion. Boston changed its runway use fee from one based
only on aircraft weight to one that combined a non-weight-based fee
and a smaller weight-based component. The fee for a small single-
engine plane increased from $25 to about $100, while the fee for a large
jumbo 747 jet decreased from $800 to less than $500. By flattening the
landing fee, Logan made it relatively more costly to land small planes,
decreasing their volume. This allowed it to more easily accommodate
the larger planes that carry more passengers. The result was that Logan
airport reduced delayed landings from 30 percent to 14 percent in less
than 4 months. Despite a reduction in congestion, the new landing
fee structure abruptly ended when the program was deemed to be in
violation of the Federal Aviation Act.
The auctioning of runway access for planes may prove to be an even
more effective way to reduce congestion at airports. An auction would
award landing rights to the carrier that values the slot the most. Such
auctions have been successful in other contexts such as to allocate radio
waves while still accommodating smaller local and public radio stations.

146 | Economic Report of the President

 Conclusion
The transportation industry relies overwhelmingly on petroleum for fuel.
In spite of its reliance, the market largely functions as it should; while trans-
portation is particularly unresponsive to changes in oil prices in the short run
due to the lack of readily available substitutes, it does eventually respond.
Also, the price reflects the costs to the firm of producing the oil and the bene-
fits to drivers from consuming the oil. That said, the use of oil by the
transportation and other sectors generates costs to national security and the
environment that users typically do not take into account. Likewise, the full
costs of congestion are not taken into account by individual users when they
drive, since roadway use is not priced by the market. Carefully crafted poli-
cies could help address these costs but care should be taken as government
action itself imposes inefficiencies.

Chapter 6 | 147