Transportation exists to help people and businesses overcome the distance between places (e.g., work and home, factory and store, store and home). Two concepts, mobility and accessibility, are most often used to measure the success of the transportation system. Mobility measurements focus on how often and far people and goods travel. Accessibility is a measure of the relative ease with which people and businesses can reach a variety of locations. Mobility and access are often positively related, but not always. For instance, less travel (lower mobility) might be the result of better access in cases where opportunities are located nearby. Many factors affect mobility and access, including the availability and cost of transportation and the infrastructure in place to facilitate it, population growth and economic fluctuations, and the knowledge of and ability to apply logistical options (particularly for businesses).
Both mobility and accessibility were affected by the terrorist attacks on the United States in September 2001. At the time this report was prepared, only the immediate and short-term impacts were known and, in many cases, only anecdotally. Air travel and some freight movements within and to and from the United States were halted entirely for several days after the attacks. By the end of the year, air travel had not returned to its previous activity level. General aviation was shut down for a longer period, and in early 2002 a few airfields in the vicinity of Washington, DC, were still closed. Maritime shipments were slowed because of a new Coast Guard policy of boarding all foreign ships to check manifests prior to their arrival in U.S. ports. Increased inspection at land crossing points has delayed shipments as well. Intercity train and bus travel rose in the immediate aftermath, but whether this will result in a fundamental shift in mode choices is unknown.
Even in the months before September 2001, transportation had been affected by slower economic growth. Prior to that downturn and during a prolonged period of expansion, however, both passenger travel and goods movement were increasing. About 4.8 trillion passenger-miles of travel were supported by the system in 2000, an annual increase of 2.3 percent since 1990. In addition, there were over 3.8 trillion ton-miles of domestic freight shipped in 1999, representing an annual growth rate of 2.0 percent since 1990.
Increases in population, numbers of workers, vehicle availability, and disposable personal income are among the factors that contribute to passenger travel growth. This growth can be seen, for instance, in international travel. Between 1990 and 2000, the number of U.S. residents traveling out of the country rose 36 percent. Growth is also evident when measured by mode.
Highway passenger travel continues to grow, with travel in light trucks (including minivans, pickups, and sport utility vehicles) posting the largest increases. The light truck share of passenger-miles of travel grew from 14 percent in 1975 to 32 percent in 2000. Despite some gains for the transit mode, the number of people driving to work alone continued its upward trend along with the distance traveled. Accessibility measures show growth, as well: the number of household vehicles, for instance, has risen to equal the number of licensed drivers. Nevertheless, there were 9.5 million households without a car in 1999. By 2000, just over 83 percent of the nationwide fleet of transit buses were equipped with ramps or lifts to provide access to the disabled. However, only about 34 percent of heavy-rail transit stations were accessible by 2000.
Congestion on the highways and in the skies slows traffic and creates a drag on the nation's economic productivity. On the highways, hours of delay per person more than tripled between 1982 and 1999, with people in the largest metropolitan areas suffering from the worst congestion. Each person in the largest metropolitan areas lost an average of 41 hours in 1999. Flight delays tend to vary from year to year making comparisons difficult. In 2001, 22 percent of flights by major U.S. air carriers were delayed, canceled, or diverted. Causes of congestion in the air and on the highways show some similarities: system capacity that is not keeping pace with increasing volumes and delays caused by inclement weather. Because data are not regularly collected for waterborne transportation, measures of the extent of congestion for this mode are not available.
Economic activity is a key factor affecting freight movement. So, too, are changes in business logistics, such as the location of distribution centers at greater distances from consumers and the wide use of just-in-time manufacturing. Air carrier and intercity trucking ton-miles are increasing at a faster rate than the other modes. Compared with other freight modes, air is used more often to move higher value commodities over longer distances. Despite the rapid growth of goods movement by air, however, most freight (measured in tons) is moved by trucks.
All modes of transportation continued to show growth in passenger-miles of travel (pmt) through 2000. Light trucks (pickups, minivans, and sport utility vehicles) posted the biggest gains, increasing its share of pmt from 14 percent to 32 percent over the 1975 to 2000 period. In absolute terms, passenger travel in light trucks grew from 363 million miles in 1975 to 1.5 trillion miles in 2000. The passenger car share of pmt declined from 76 percent in 1975 to 53 percent in 2000. Air travel also increased its share from 5 percent to 11 percent. Overall, pmt, excluding miles traveled in heavy trucks, grew from about 2.6 trillion in 1975 to almost 4.8 trillion in 2000 (figure 1). On a per capita basis, people traveled 17,000 miles in 2000 compared with 11,900 in 1975 .
Several factors contributed to the continued growth in pmt (figure 2). The resident population, for example, increased by nearly 66 million people, a rise of 31 percent between 1975 and 2000. Moreover, the number of people in the civilian labor force, most of whom commute to work, grew almost twice as fast as the population over the same period. People also have more money to spend on transportation, particularly for automobiles and air travel. Disposable personal income per capita rose from $14,393 in 1975 to $23,640 in 2000 (in chained 1996 dollars) .
An increasing number of people can now afford to buy vehicles and travel services, especially since the cost of the most widely used kinds of transportation-travel in cars and planes-fell in real terms. For example, the inflation-adjusted average airfare for domestic scheduled service declined from $174 in 1975 to $110 in 1995 and has stayed at that level through 1999 (measured in chained 1996 dollars) . Despite recent fluctuations, gasoline prices, too, have been at historically low levels for much of the past 15 years . Intercity rail fares also decreased between 1975 and 1999, but average intercity bus fares increased more than inflation during this period. Average bus fares went from $17 to $23 between 1975 and 1999 (in chained 1996 dollars) . Rising bus fares tend to affect individuals with lower incomes more than people at higher income levels.
See box for Estimating Passenger-Miles of Travel.
1. American Petroleum Institute. "How Much We Pay for Gasoline: 1999-April 2000 Review," May 2000, available at http://www.api.org/pasp/biggas.pdf, as of Sept. 9, 2000.
2. U.S. Department of Commerce, U.S. Census Bureau, Statistical Abstract of the United States, 2001 (Washington, DC: 2002).
3. U.S. Department of Transportation, Bureau of Transportation Statistics, National Transportation Statistics 2001 (Washington, DC: 2002), table 3-15b.
With increases in both population and individual travel, highway usage has risen substantially. Annual vehicle-miles of travel (vmt) in the United States rose by nearly 30 percent to 2.8 trillion miles between 1990 and 2000, an annual increase of 2.5 percent. Vmt per capita rose by just over 13 percent during the same period, an annual increase of 1.3 percent.
The most heavily populated states, California, Texas, Florida, and New York, are the most heavily traveled. However, Wyoming, the least populated state, had the highest vmt per capita in 2000 at 16,400, followed by Georgia, Alabama, Oklahoma, and New Mexico at over 12,500. The District of Columbia and New York had the lowest vmt per capita at just under 7,000. The percentage change in vmt per capita between 1990 and 2000 ranged from a 32 percent increase in Mississippi to a 3 percent decline in Hawaii, with 12 states showing an increase of at least 20 percent over the 10-year period (see map).
In recent years, the makeup of the U.S. vehicle fleet changed as well, altering the share of vmt by vehicle type (figure 1). While the share of total vmt by buses and single-unit and combination trucks has remained relatively constant, the increasing popularity of sport utility vehicles and other light trucks in recent years has resulted in a shift in the percentage of total vmt from automobiles to light trucks. Although still the dominant vehicle type in terms of vmt, the share of automobile vmt declined from 66 percent of total vmt to 58 percent between 1990 and 2000. Over the same period, the percentage of total vmt by light trucks (a classification including minivans, pickup trucks, and sport utility vehicles) rose to 34 percent of total vmt .
See figure 2 for Increases in Passenger-Miles of Travel (PMT) and Factors Affecting Travel Demand: 1975-2000.
See box for The Highway Performance Monitoring System.
1. U.S. Department of Transportation, Federal Highway Administration, Highway Statistics (Washington, DC: Annual issues).
Overnight travel between the United States and foreign countries for both business and pleasure grew during the 1990s (figure 1 and figure 2). However, the terrorist attacks of September 2001 and the preceding economic slowdown caused a decline in late 2001 and early 2002 in land border crossings and air travel (box 1).
International travel data does not take into account people staying for less than one night (box 2). Still, the decade-long growth in overnight travel has implications for the infrastructure at America's borders (including airports and land border crossings) and the demand on transportation infrastructure by foreign nationals while they are in the country. There are also economic implications related to travel spending.
Factors that have contributed to growth include the globalization of the production of goods and services, lower priced air transportation, economic growth, and rising incomes in many parts of the world. The United States received 7 percent of the nearly 699 million international overnight visitors in 2000. Expenditures by these visitors totaled $476 billion worldwide, with 17 percent spent in the United States [1, 2].
In 2000, a record 51 million overnight international visitors traveled to the United States. Nearly three-quarters of them were from five countries: Canada, Mexico, Japan, the United Kingdom, and Germany (table 1). The number of visitors from overseas (all countries except Canada and Mexico) has risen in the past few years (figure 1). Canadian travel to the United States has been increasing since 1998, after a gradual decline through most of the decade.
In 2000, U.S. residents made more than 60 million international trips. Major destinations were Mexico, Canada, and the United Kingdom (table 2). International travel by U.S. residents between 1990 and 2000 grew by more than 36 percent, with travel overseas growing the fastest (figure 2).
1. U.S. Department of Transportation, Bureau of Transportation Statistics, U.S. International Travel and Transportation Trends (Washington, DC: 2002).
2. World Tourism Organization, Organization Tourism Market Trends (Madrid, Spain: Sept. 25, 2001).
Over 290 million people entered the United States at crossing points on the U.S.-Mexico border in 2000, more than triple the 95 million entering on the U.S.-Canada border (table 1). Most people traveled across the border in personal vehicles, although a large number of people entered the United States from Mexico on foot. El Paso, Texas, and San Ysidro, California (near San Diego), were the top vehicle crossing points. On the Canadian border, the top crossing points were Detroit, Michigan, and Buffalo-Niagara Falls, New York (table 2).
Although more than 800 airports in the United States provided some form of air passenger service to the public in 2000, most enplanements (i.e., passenger boardings) occur at a relatively small number of airports. In 2000, for instance, 84 percent of all U.S. air passengers enplaned at 50 airports (table 1). Ten airports alone accounted for about one-third of all 2000 enplanements.
Air travel became more affordable during the 1990s leading to a general increase in passenger traffic at U.S. airports. Between 1990 and 2000, enplanements at all airports grew by nearly 50 percent from 439 million to 639 million [1, 2]. Some airports grew much faster than the average and/or experienced very large growth in the number of passengers boarded. Enplanements doubled at six major airports (of the top 50 in 2000) between 1990 and 2000: Greater Cincinnati, Sacramento International, Portland International, McCarren International (Las Vegas), George Bush Intercontinental (Houston), and Baltimore-Washington International. Moreover, 5 major airports increased boardings by 7 million or more over this period: Hartsfield International (Atlanta), McCarren International, George Bush Intercontinental, Minneapolis-St. Paul International, and Detroit Metropolitan Wayne County.
Factors leading to a rapid rise in enplanements at specific airports include: location in or near a rapidly growing metropolitan area (e.g., Las Vegas, Phoenix, and Atlanta) or major tourist destination (e.g., Las Vegas and Orlando), serving as a hub for a major commercial airline (e.g., Cincinnati and Houston), and serving a fast growing low-fare airline such as Southwest Airlines (e.g., Baltimore-Washington International).
1. U.S. Department of Transportation, Bureau of Transportation Statistics, Airport Activity Statistics of Certificated Air Carriers: Summary Tables, Twelve Months Ending December 31, 2000 (Washington, DC: 2001).
2. U.S. Department of Transportation, Federal Aviation Administration and Research and Special Programs Administration, Airport Activity Statistics of Certificated Route Air Carriers, Twelve Months Ending December 31, 1990 (Washington, DC: 1991).
Accessibility for all Americans is a fundamental goal of public transit services. This includes access to bus and demand responsive transit, rail transit (heavy, commuter, and light), and other transit modes (trolley and ferry). The Americans with Disabilities Act (ADA), in fact, requires public transit services (fleets and facilities) to be accessible to persons with special needs.
The nationwide fleet of ADA lift- or ramp-equipped transit buses increased from 52.2 percent of the fleet in 1993 to 83.6 percent of the fleet in 2000 (table 1). While greater compliance with ADA requirements can be seen from 1993 to 2000, the rate of compliance has differed among types of buses (figure 1). The small bus fleet had the highest level of compliance in 1994 and articulated buses, the lowest compliance. In 2000, the small bus fleet continued to have the highest compliance, while large buses had the lowest compliance .
Rail transit infrastructure consists of track and stations. At a maximum service level, 14 heavy-rail transit agencies operate 8,245 vehicles over 2,163 miles of track. Among the 997 rail stations in 1997 serving heavy rail, 25.7 percent (256 stations) were ADA accessible. Between 1997 and 2000, the total number of heavy-rail stations increased by only 12 (to 1,009), while the number of ADA accessible stations increased 32.8 percent (to 340). Still, by 2000, ADA accessible stations equaled only 33.7 percent of the total number of stations, leaving a balance of 66.3 percent inaccessible. The level of compliance by heavy-rail transit agencies ranges from 0 percent to 100 percent (table 2). The New York City Transit Authority system, for instance, with 46 percent of the total number of stations, has an accessibility rate of just 8 percent. Excluding these stations from the total increases the overall accessibility rate for the rest of the heavy-rail system to 55.3 percent.
1. U.S. Department of Transportation, Federal Transit Administration, 2000 National Transit Summaries and Trends, available at http://www.ntdprogram.com, as of Feb. 15, 2002.
Nearly 9 out of 10 workers in 2000 traveled to work by car, truck, or van (table 1), and most of those driving to work drove alone . These Census 2000 Supplementary Survey data are not directly comparable with 1990 decennial census data for several reasons, including the fact that the former does not cover the group quarters population,1 perhaps understating some categories, such as walking. However, the national trends revealed by a comparison of the two data sets are consistent with other evidence of changes in how people get to work . In particular, the census data show that the share of workers driving alone to work increased and carpooling decreased between 1990 and 2000, while the share of workers using public transportation remained about the same (figure 1).
1 The group quarters population consists of people not living in households. It includes, for instance, those who are institutionalized and people who live in group housing, such as college dormitories, military quarters, and group homes.
1. U.S. Department of Commerce, U.S. Census Bureau, Census 2000 Supplementary Survey for the United States, available at http://factfinder.census.gov/home/en/C2SS.html, as of Mar. 27, 2002.
2. U.S. Department of Transportation, Bureau of Transportation Statistics, Transportation Statistics Annual Report 2000 (Washington, DC: 2001).
Transit ridership grew steadily between 1995 and 2000 to reach 9.4 billion unlinked1 trips in 2000 , an increase of 21 percent. The number of trips made in 2000 represents a 3.9 percent annual increase since 1995 and the highest ridership in more than 40 years . Rail transit ridership posted particularly strong growth (figure 1). Between 1995 and 2000, heavy rail grew 32 percent, followed by commuter rail at 20 percent, and light rail at 17 percent. Bus ridership dipped in the mid-1990s but by 2000 had risen back to the level of ridership held in 1990. Most transit trips are still taken by bus [1, 2].
1 Each time a passenger boards a vehicle it is counted as an unlinked trip, regardless of the number of vehicles the passenger must board to travel from origin to destination.
1. American Public Transit Association, "APTA Transit Ridership Report," available at http://www.apta.com/stats/ridershp/riderep/history.pdf, as of Sept. 7, 2001.
2. ____. Public Transportation Fact Book 2001 (Washington, DC: 2001).
3. ____. "Public Transportation Ridership Tops 9.4 Billion in 2000," available at http://www.apta.com/news/releases/2000rides.htm, as of Sept. 7, 2001.
Because of improvements in vehicle reliability and longevity and rising incomes, more people now own a motor vehicle than 10 years ago. However, about 9.5 million households-representing 9 percent of all households-were without a car, van, or truck in 1999 . This is down from 10.1 million households (11 percent of all households) in 1989 .
Black, Hispanic, poor, and elderly households are more likely to be without a car, van, or truck than the population as a whole, despite relatively large increases in vehicle ownership among these groups (figure 1). Not surprisingly, poor households are the least likely to have vehicles. The number of households below the poverty level without a vehicle dropped 10 percent, from 37 percent in 1989 to 27 percent in 1999.
The geographic location of a household also affects vehicle ownership. For instance, households in central city urban areas are less likely to own motor vehicles than households in the suburbs, urban areas outside a metropolitan statistical area, or rural areas (figure 2). Similarly, when data are aggregated on a regional basis, the heavily urban Northeast has the highest share of households without vehicles (figure 3).
See box for "Census Island," USA.
1. U.S. Department of Housing and Urban Development and U.S. Department of Commerce, U.S. Census Bureau, American Housing Survey for the United States: 1999, H150/99 (Washington, DC: 2000).
2. _____. American Housing Survey for the United States: 1989, H150/89 (Washington, DC: 1990).
Being stuck in traffic is a source of frustration for many travelers, particularly commuters, but the impacts go far beyond those individuals immediately affected. By wasting people's time, increasing the time it takes to transport goods, and causing missed meetings and appointments, highway congestion is a drag on economic productivity. Congestion is also an environmental concern. Extra fuel is consumed by cars traveling under these conditions because of increased acceleration, deceleration, and idling. Greater fuel consumption leads to higher emissions of greenhouse gases and may raise the level of other air pollutants.
The Texas Transportation Institute (TTI) studies 68 metropolitan areas in order to estimate congestion and some of its impacts in the United States. TTI found that between 1982 and 1999 congestion measured by average annual delay per person increased in all areas (see map). Overall in the study areas, average annual delay per person has more than tripled during the 17-year period, rising from 11 hours per person in 1982 to 36 hours in 1999 (table 1). Furthermore, drivers in the largest metropolitan areas (with a population of over 3 million) experienced the worst congestion (41 hours per person on average in 1999), and those in small metropolitan areas (population of 500,000 or less) the least (10 hours a year per person) (figure 1).
Delayed or canceled commercial airline flights cost consumers in many unmeasured ways, including lost personal time, missed meetings, and increased anxiety and stress. Delay also costs the airlines. The Federal Aviation Administration (FAA) estimates that commercial aviation delays cost airlines over $3 billion annually and projected in 2000 that delays throughout the system would continue to increase as the demand for air travel rose . Both FAA and the airlines consider that improvements in air traffic control should mitigate some flight delay problems. In addition, FAA and the industry, under the FAA's National Airspace System (NAS) Operational Evolution Plan, have begun to implement ways to reduce delays in the national aviation system attributable to weather, increasing flight volume, and limited system capacity .
Both the Bureau of Transportation Statistics (BTS) and FAA track airline delays. According to BTS, a flight is counted as an "on-time departure" if the aircraft leaves the airport gate less than 15 minutes after its scheduled departure time, regardless of the time the aircraft actually lifts off from the runway. Also, BTS counts an arriving flight as "on time" if it arrives less than 15 minutes after its scheduled gate arrival time.
Unlike BTS, which tracks air carrier performance, FAA tracks delays in terms of how well the air traffic control system performs . Tracking begins once a flight is under FAA air traffic control (i.e., after the pilot's request to taxi out to the runway). As such, an aircraft could wait an hour or more at the gate before requesting clearance to taxi. Once under air traffic control, as long as the aircraft took off within 15 minutes of the airport's standard taxi-out time, FAA considers the flight departed on time. .
About one-quarter of flights by major U.S. air carriers were delayed between 1996 and 2001, according to BTS data (figure 1). In 2001, 22 percent of flights were delayed, canceled, or diverted, down from a 10-year high of 27 percent in 2000 .
Most delays take place while a plane is on the ground, although the actual cause of a delay may occur elsewhere. Poor weather is the most common cause of delays (figure 2). The growth in flight volume was a major contributor to delays and cancellations during the 1990s. The total number of flight operations1 at the nation's airports increased by 7 percent, from 63.0 million to 67.7 million flights, between 1992 and 20002 . Delayed flights rose from less than one in five flights to more than one in four flights during the same period .
A slowing economy contributed to a 3 percent reduction in total flight operations between August 2000 and August 2001 . The demand for flights also dropped considerably following the terrorist attacks in September 2001 (box 1). The reduction in flight volume during 2001 contributed to a 5 percent decrease in flight delays compared with 2000 .
There is much debate about the role of airline scheduling in causing delays. The hub and spoke systems used by the major airlines concentrate flights into the hub airports. The worst delays tend to be at peak travel times during the day and at certain times of the year (e.g., holidays and the summer months) when travel volume is heavier. When heavy volume is combined with bad weather between a hub airport and its spokes, the ripple effect can cause delays at dozens of other airports  (table 1).
In August 2000, the U.S. Department of Transportation (DOT) created a task force comprising a cross-section of aviation stakeholders, including representatives from airlines, consumer groups, labor unions, and airport operators, to examine the reasons for flight delays and develop recommendations on how to modify airline on-time reporting. Currently, the on-time information that the 10 largest U.S. passenger carriers are required to submit to BTS3 identifies only the frequency and duration of flight delays and cancellations, not the cause. As a preliminary step to expanding rules for air carrier on-time reporting, the task force implemented a pilot test program for reporting causes of flight delays . DOT also developed a notice of proposed rulemaking that would modify the current regulations governing the submission of on-time flight performance reports to require the reporting of causal information relative to flight delays and cancellations.
1Flight operations, as reported by FAA, include takeoffs and landings by all types of aircraft (commercial and general aviation) at approximately 3,400 domestic airports.
2 Comparable data are not available for 1990 and 1991.
3 American Eagle was added to the list of carriers that must report on-time data and began reporting in January 2001. This carrier is excluded from the data and figures in this report to retain comparability with previous years.
1. Mead, K.M., Inspector General, U.S. Department of Transportation, "Flight Delays and Cancellations," statement before the Committee on Commerce, Science, and Transportation, United States Senate, Sept. 14, 2000.
2. U.S. Department of Transportation, Bureau of Transportation Statistics, Office of Airline Information, personal communications, November 2000-November 2002.
3. U.S. Department of Transportation, Federal Aviation Administration, NAS Operational Evolution Plan, available at http://www.faa.gov/programs/oep/OMDEX.htm, as of April 2002.
4. _____. OPSNET database.
Freight movements grew significantly over the past quarter century despite a general trend in the economy toward services and high-value, low-weight products. Between 1975 and 1999, domestic freight ton-miles increased 67 percent, from 2.3 trillion to 3.8 trillion, with air carriers and intercity trucking growing faster than the other modes (figure 1). Despite the decline in the maritime mode since 1980, attributable to the decline in Alaskan crude oil shipments, water transportation still accounted for 656 billion ton-miles in 1999.
Population growth and economic activity are some of the factors that determine freight demand; increases in both mean a greater volume of goods produced and consumed and thus more freight moved (figure 2). Between 1975 and 1999, the resident population rose by 57 million, an increase of 26 percent, while the gross domestic product more than doubled from $4 trillion to $8.9 trillion (in inflation-adjusted chained 1996 dollars). The growth in freight ton-miles was slower than the growth in economic activity during this period but outpaced the increase in population.
As economic activity expanded, particularly in the 1990s, changes in what, where, and how goods were produced affected freight demand and contributed to the increase in total ton-miles. The composition of goods produced also changed as the economy shifted toward more services and high-value, low-weight products. This shift can be measured by the ratio of ton-miles per dollar of Gross Domestic Product (GDP), which has declined since 1975. This decline suggests that, as the economy becomes more service-based, it is also becoming less freight transportation intensive. For instance, it takes more freight ton-miles to produce $1,000 worth of steel than it does to produce $1,000 worth of cellular phones. Today, even traditional products, such as automobiles, are made from lighter, but often more expensive, materials such as engineered plastics.
However, freight ton-miles per capita rose more than 30 percent, from about 10,600 in 1975 to 14,000 in 1999. As economic growth has accelerated, disposable personal income per capita has increased and individual purchasing power risen. Businesses have responded by shipping more freight per resident population.
The manufacture, assembly, and distribution of goods continue to change as components of products are produced in facilities located thousands of miles apart, some halfway around the globe. Today, many businesses manage worldwide production and distribution systems, increasing global trade in goods and the demand for freight transportation. Changes in where goods are produced can directly increase total ton-miles and change the average length of haul of shipments. Such changes also affect freight mode choice, with more commodities being shipped by multiple modes as distances increase. This worldwide spatial distribution of production activities and trade impacts transportation requirements in the United States. For example, expanding trade with the Pacific Rim continues to make West Coast container ports more dominant than East Coast ports and poses challenging landside and intermodal access demands.
See box for Freight Analysis Framework Tool.
During the past decade, the U.S. domestic air cargo industry moved increasing amounts of freight while providing speedy, reliable, and safe freight services to support business activity and household package delivery in the United States. Between 1991 and 2000, domestic air cargo tonnage more than doubled, from 5 million tons to 13 million tons, an annual growth rate of 11 percent (table 1). Measured in revenue ton-miles, air freight grew 6 percent annually between 1991 and 2000.
Air freight remains a vital link that enables quick access by firms and households to goods produced throughout the United States and the world. Since 1991, the air freight .ton-miles per capita has increased over 50 percent, rising to almost 53 ton-miles per person per year by 2000. At the same time, miles per ton-mile declined by over 35 percent (table 1). This decline in average air cargo flight distance may be a result of changes in the marketplace, especially during the second half of the decade when enplaned revenue tons increased at a much greater rate than revenue ton-miles. Contributing factors include changes in the competitive relationship between air and surface modes, increasing the attractiveness of air for short-haul shipments; creation of new markets for short-haul air shipments because of fundamental changes in the U.S. economy, as discussed below; and a maturing of the air cargo system making it more competitive for short-haul shipments.
Strong growth in the U.S. economy during the 1990s, evolving production and distribution systems, and interest in electronic commerce have spurred growth in air cargo. Air freight has outpaced increases in real Gross Domestic Product (GDP) and U.S. resident population (figure 1). Furthermore, air freight grew significantly in line with a general trend in the economy toward services and high-value products. This is evidenced by the trends in revenue ton-miles per dollar of real GDP.
Air cargo continues to be impacted by a shift toward services and new methods of product manufacturing, production, and distribution. While scheduled air service rose in response to demands to move air cargo quickly during the past decade, nonscheduled service grew much faster, transporting only 3 percent of domestic enplaned revenue tonnage in 1991 but 40 percent in 2000 (figure 2). Nonscheduled service gives businesses greater flexibility to move cargo fast and efficiently on short notice and may provide a quicker response to just-in-time business demands.
Unused freight ton-miles, a measure of excess capacity in the air cargo industry, steadily grew between 1991 and 2000 (figure 3). In 2000, 27 billion ton-miles of air cargo capacity were unused, accounting for nearly 65 percent of the available freight ton-miles. The relative proportion of excess capacity, however, remained stable during the decade. The average air cargo load factor also remained stable at around 35 percent despite an overall increase in available capacity to nearly 42 billion ton-miles in 2000.
U.S. container trade increased nearly 7 percent from 1999 to 2000. This trade is concentrated in 25 ports and has become more concentrated in the last 4 years in 10 U.S. container ports (table 1).
Prior to the economic slowdown that began in late 2000, demand for U.S. exports was not keeping pace with U.S. consumer demand for imports. Accordingly, the balance of international container trade (i.e., the volume of U.S. containerized exports compared with containerized imports) shifted in favor of U.S. imports, particularly in recent years. Between 1993 and 1997, the balance of U.S. international container trade was less than 1 million 20-foot equivalent units (TEUs) per year (figure 1). By 2000, this gap had widened to a difference of over 4 million TEUs.
Of the top 10 U.S. container ports, the Port of Los Angeles has the deepest maintained channel depth, followed by the Port of Long Beach. The largest container vessel serving U.S. ports has a draft of 45 feet . Out of 441 containerships on order in world shipyards as of June 2001, 98 will have capacities greater than 5,000 TEUs . Because there are few U.S. ports with channel depths sufficient for such containerships, the ports may evolve into a system whereby a main port (i.e., hub) feeds cargoes to a network of ports using smaller draft ships.
1. Journal of Commerce, "Special Report: Top 50 Container Lines," JOC Week, vol. 2, no. 34, Aug. 27, 2001-Sept. 2, 2001.
2. Lloyd's Register of Shipping, World Shipbuilding Statistics (London, England: June 2001).
Intermodal movement of freight-shipments transported by multiple modes-has grown sharply in recent decades. Rail shipments with an intermodal component have steadily increased, containers transported by ship that typically move intermodally have largely replaced break bulk shipments, and almost all air freight shipments travel via truck at some point between their origin and destination. As the movement of intermodal freight through airports and seaports continues to rise, the potential for bottlenecks exists. These bottlenecks are likely to occur where the volume of freight moving through a facility or area exceeds the capacity of the transportation system to operate without significant and costly delays.
Transportation planners can use analytical tools to help them identify potential bottlenecks and assess the status of current intermodal freight movements. Analysis using a tool developed by the U.S. Department of Transportation (see box) shows that in 1999, Memphis, Los Angles, and Newark International Airports moved the largest tonnage of freight in the country. Among the top 40 airports, George Bush Intercontinental Airport in Houston, Texas, was ranked first by the freight-to-capacity ratio1 (table 1). This suggests that although Houston's airport ranks 21st in tonnage moved, its throughput is nearly three times the current infrastructure capacity. However, having a higher than average freight-to-capacity ratio on an annual basis does not necessarily result in bottlenecks since freight shipments may vary daily and move during off-peak hours.
Highway traffic volume and delay around intermodal facilities caused by both freight and passenger vehicles also directly impacts the effectiveness of transferring freight between modes. When freight airports are ranked by the average annual daily traffic (AADT) count within a five-mile radius, the most congested airports are San Francisco, Chicago O'Hare, and Oakland, California. These three airports are also at the top of the list when ranked by traffic delays per lane-mile within five miles of the airport facilities.
The intensity of use and the potential for intermodal bottlenecks for the top U.S. seaports based on the tons of throughput can also be displayed (table 2). The Port of Long Beach, California, ranks first in AADT per lane-mile and delay per lane-mile, followed by ports of Oakland and Richmond, California. On a freight-to-capacity ratio, however, Baton Rouge, Louisiana, and Port Arthur, Texas, are the highest rated ports.
1This ratio compares the volume of freight transported to the relative capacity of the transportation infrastructure in handling that level of freight. It divides freight traffic on the transportation network by a measure of nominal flow representing a reasonable annual traffic level for facilities of its kind. Assigned flows in excess of nominal flow do not automatically imply a capacity problem, merely that flows are above average for that type of facility.
Railroad intermodal traffic-moving a trailer or container on a rail flatcar-has grown faster than any other segment of the railroad industry, tripling since 1980. Containers-on-flatcars (COFC) have led this growth in recent years with the increase of doublestack container service fueled by higher levels of international trade from the Pacific Rim countries.
Intermodal traffic grew from 8.1 million COFCs and trailer-on-flatcars (TOFCs) in 1996 to 9.2 million in 2000, an increase of almost 14 percent (figure 1). During the same period, COFC shipments increased over 31 percent while TOFC shipments dipped about 12 percent. Furthermore, the number of railroad flatcars used for intermodal transportation declined overall, indicating an improvement in railroad equipment utilization and increased productivity (figure 2).
The largest flows of intermodal traffic are between California and Illinois and consist primarily of moving containers from ships to their final destinations. Overall, intermodal rail shipments move an average 1,400 miles, much longer than the average rail move of less than 900 miles . Intermodal rail equipment is used in 16 percent of all ton-miles of rail traffic and is second only to coal in ton-miles carried. The great majority of intermodal rail traffic, 69 percent, is classified as Miscellaneous Mixed Freight and is thought to involve primarily the shipment of lightweight, high-value retail goods.
1. U.S. Department of Transportation, Surface Transportation Board, Carload Waybill Sample, 1999.
For several decades, the motor carrier industry and government agencies have been investigating ways to improve commercial vehicle operations with the use of intelligent transportation systems (ITSs). Many of these systems are in use today, such as vehicle onboard safety monitoring and interstate exchange of driver credentials. Other highway ITSs, for example, real-time traffic monitoring, remotely controlled high-occupancy vehicle (HOV) access gates, and electronic toll collection, are not specific to commercial vehicle operations but can improve them.
PrePass is one ITS available for commercial vehicles in 25 states (see map). This automatic vehicle identification system uses transponders on truck windshields to enable participating commercial vehicles to bypass designated weigh stations and port-of-entry facilities. By eliminating the need to stop at highway weigh stations, PrePass improves shipping efficiency and safety for all highway users and reduces fuel consumption and vehicle maintenance by decreasing vehicle braking and acceleration.
PrePass evolved out of a project initiated by the Federal Highway Administration (FHWA) in the 1980s. A demonstration and evaluation program involving three states (Oregon, California, and Arizona); Alberta, Canada; and the trucking industry was completed in 1994. Subsequently, California and Arizona opted to continue the project and formed a nonprofit partnership (Heavy Vehicle Electronic License Plate (HELP), Inc.) with motor carriers to promote and operate PrePass. HELP, Inc. has set up over 220 operational sites in 25 states and registered over 7,900 carriers with more than 170,000 vehicles into  (figure 1).
There are other cooperative efforts between states to implement similar systems. For instance, the NORPASS system is in use in seven northwestern states. Oregon's Green Light program operates at 21 weigh stations, handling over 1,000 carriers with more than 14,000 trucks . In 1999, FHWA initiated rulemaking to establish standards and specifi-cations for the transponders used in various weigh station pre-clearance systems [3, 4]. FHWA expects to issue the rules after completing the testing of transponders.
1. Heavy Vehicle Electronic License Plate (HELP), Inc., available at http://www.prepass.com/, as of Nov. 1, 2001.
2. Oregon Department of Transportation, "Oregon Green Light," available at http://www.odot.state.or.us/trucking/its/green/light.htm/
3. _____. "U.S. DOT Seeks Transponder Standards to Ensure Interoperability," available at http://www.odot.state.or.us/trucking/its/transponders.htm/, as of Nov. 1, 2001.
4. U.S. Department of Transportation, Docket Summary Information, Docket No. FHWA-1999-5844, available at http://dms.dot.gov/, as of Nov. 1, 2001.