Chapter Four - Transportation Safety

The highest priority of the U.S. Department of Transportation (DOT) is to "promote public health and safety by working toward the elimination of transportation-related deaths, injuries, and property damage" (USDOT NHTSA 1997). Although great progress has been made in the last few decades, transportation crashes remain the leading cause of accidental deaths and injuries in the United States, claiming tens of thousands of lives, and injuring millions of people each year. Most of the transportation deaths (about 95 percent) and injuries (an even higher percentage) occur on the nation's highways.

Although many safety issues are specific to a particular mode, others cut across modes. This chapter briefly discusses mode-by-mode trends, but concentrates on cross-modal issues.¹ Following a review of long-term transportation safety trends, the chapter discusses four special concerns-safe transportation to and from schools, pedestrian and bicycle safety, security of transportation systems against crime and terrorism, and international trends in road safety-as well as data needs.

Safely transporting students to and from schools is of particular interest to the public. Congress, in the Transportation Equity Act for the 21st Century (Public Law 105-178), known as TEA-21, mandated DOT to undertake a special study of safety issues in transporting school children using various modes, which is due in 2001.

Walking and biking are often promoted as healthy, nonpolluting forms of transportation that could help alleviate congestion. However, walkers and bicyclists account for roughly 15 percent of fatalities involving motor vehicles, and their safety is an important concern.

In a pervasively integrated world economy, with more Americans traveling abroad for business and pleasure, international trends in transportation safety and security are of interest. Transportation, particularly air travel and mass transit can be targets of terrorists: for example, the gas attack on the Tokyo subway. Air travelers now regard precautions against terrorism as a routine part of air travel. In addition, this chapter briefly highlights road safety issues in other countries, as U.S. citizens traveling abroad who use taxis, buses, or rental cars may encounter traffic safety conditions that can differ from those in the United States.

LONG-TERM TRENDS

Table 4-1 shows historical trends in absolute numbers of fatalities, injuries, and crashes/accidents for all modes. The data show that in the last three decades, transportation fatalities have decreased for most modes. The figures for rail and transit include fatalities and injuries arising from certain types of incidents involving no moving vehicle (e.g., falls on transit property or fires in railroad repair sheds). Thus, the data in table 4-1 for transit and rail include many nonpassengers, making comparisons with other modes difficult.

Table 4-2 shows the distribution of fatalities in 1997. Occupants of cars and light trucks accounted for nearly three-quarters of the fatalities, and motorcyclists slightly under 5 percent, while pedestrians and pedalcyclists hit by motor vehicles accounted for just under 14 percent of the fatalities. Recreational boating and general aviation accounted for over 3 percent of the fatalities-more than all of the listed commercial passenger modes and freight trains combined.

Table 4-3 provides historical detail for the highway mode, which accounts for most transportation fatalities and injuries. As shown, there are fewer fatalities involving motor vehicles today than in the mid-1970s or early-1980s, even though miles of travel have increased. The decline is evident across most highway submodes, with the conspicuous exception of light trucks (e.g., sport utility vehicles, vans, and pickup trucks). The number of such vehicles increased dramatically over the period; they now account for about one-third of the motor vehicle fleet. The implications of these vehicles for highway safety is receiving considerable scrutiny. For example, sport utility vehicles have the highest rollover involvement rate by far of any vehicles in fatal crashes-36 percent compared with 15 percent for passenger cars (USDOT NHTSA 1998b).

Considerable concern also exists about the increasing number of crashes involving light trucks and passenger cars. A study sponsored by the National Highway Traffic Safety Administration (NHTSA) concluded that approximately twice as many car drivers would die in collisions with light trucks than in similar collisions involving cars of the same weight as the light truck (Joksch 1998, xi).

For systematic analyses of transportation safety trends, it is customary to calculate safety rates by dividing the absolute numbers of fatalities (or other adverse outcome) by some measure of activity, such as number of trips, miles traveled, or hours of vehicle operation. The logic for doing this is straightforward; people are exposed to transportation risks primarily when they travel. Hence, everything else remaining the same, the more travel there is, the more risk people incur. Thus, it can be misleading to use absolute numbers of, say, fatalities to compare the safety of a given mode for two different years, since any change in fatalities might be explained solely by a change in the amount of transportation activity.

Figure 4-1 shows fatality rates for selected modes for a time period of about two decades. Clear improvement is apparent for several modes and submodes. While it is valid to compare fatality rates in two different years for the same mode, comparisons between modes should be done only with great caution. As seen in figure 4-1, the denominators used are not the same for all modes. For highway and rail, exposure to risk is approximately proportional to distance traveled, hence the use of vehicle-miles as the denominator. For aviation, the greatest proportion of crashes occur during takeoff and landing; hence risk is approximately proportional to the number of operations (measured as departures). Data availability for general aviation is limited, so hours flown is used instead. For rail, it can be argued that the analog of highway vehicle-miles is railcar-miles rather than train-miles, but lack of railcar data necessitates the use of train-miles.

Considerable confidence can be placed in comparisons of different highway submodes. As shown in figure 4-1, occupants of passenger cars and light trucks have much higher fatality rates than occupants of large trucks and buses. Motorcycle riders have fatality rates more than an order of magnitude greater than for other highway submodes. A large number of factors influence differences in fatality rates. For example, the greater size and mass of large trucks and buses serves to protect the occupants of these vehicles in crashes with smaller vehicles or less massive objects. These weight differences explain why most of the people killed in crashes involving large trucks are in other vehicles. In 1997, 4,871 large trucks were involved in fatal crashes. While 717 occupants of large trucks died in these crashes, 4,638 other people died in these collisions (USDOT NHTSA 1998c).

There are various possible explanations for decreasing fatality rates for all modes. For highway modes, promotion of safety belt, child safety seat, and motorcycle helmet usage, and measures to discourage drunk driving have all had a beneficial effect. For example, the portion of drivers in fatal crashes who were intoxicated (blood alcohol concentrations of 0.10 grams per deciliter) declined from 30 percent in 1982 to 17.8 percent in 1997 (USDOT NHTSA 1998b, p. 34, table 15). Box 4-1 discusses safety belt and air bag use. Some of the decrease in transportation fatalities may be a consequence of better and prompter medical attention for victims of transportation crashes and accidents.

TRANSPORTATION OF SCHOOL CHILDREN

Tens of millions of children travel to and from school each day during the school year and protecting them is an important public concern. School bus safety has received the most notice, but other issues-the safety and security of children transported in cars, by transit, or while walking or biking to school-are also key.

DOT's 1995 Nationwide Personal Transportation Survey (NPTS) indicates that the leading mode of school transportation is the privately owned vehicle (POV)-private automobiles, vans, trucks, or sport utility vehicles-followed by the school bus² (see figure 4-2). In fact, POVs accounted for half of the annual person trips for school transportation. School buses provided 32 percent of annual person trips and public transportation just 3 percent. Walking made up 10 percent of annual person trips to and from school and bicycling made up 1 percent.

The majority of fatal crashes involving school-age occupants occur when children are riding in an automobile or other private vehicle. Derived from NHTSA's Fatality Analysis Reporting System (FARS) data for 1997, figure 4-3 shows the number of fatal crashes and fatalities among occupants aged 5 to 19 that occurred during the school year between the hours of 6:00 a.m. and 9:00 a.m. and 2:00 p.m. and 5:00 p.m., when most school-related trips take place. Automobiles and other private vehicles accounted for most of these incidents. Automobiles alone accounted for 57 percent of all fatal crashes involving school-age occupants and 47 percent of occupant fatalities.

A better way of assessing the safety of various modes used in school transportation is to look at occupant fatality rates. Combining 1995 data from NPTS with FARS statistics reveals that fatal injury rates are far higher for POVs than for buses: 2.6 fatalities per 100 million person-miles for POVs used in school transportation, but only 0.06 for school buses.

School Bus Safety

According to NHTSA (1998a), school buses provide one of the safest forms of transportation in the United States. In 1987 through 1997, there were about 420,000 fatal traffic crashes: of these, 1,298 (just over 0.3 percent) were school bus-related (USDOT NHTSA 1998f). Nevertheless, while the statistics show that it is far safer to travel on a school bus than in a POV, school bus safety draws great attention whenever an incident occurs.

Each day, about 440,000 public school buses transport 23.5 million children to and from school and school-related activities. These buses travel 4.3 billion miles each year (USDOT NHTSA 1998a).

From 1987 through 1996, an average of 10 school bus occupants and 25 school age pedestrians (under 19 years of age) per year died in school bus-related crashes. From 1977 through 1986, an average of 12 school bus occupants and 47 pedestrians died annually (NHTSA 1998a, 2). NHTSA also estimates that an average of about 8,500 injuries occurred each year in school buses between 1988 and 1996. Most were minor injuries, but 885 were moderate and 350 were serious to critical (USDOT NHTSA 1998a).

The debate over whether school buses should be equipped with seat belts goes back to at least 1977, when NHTSA tightened school bus safety standards. At that time, following extensive research and analysis, NHTSA instituted "compartmentalization" as the primary means of occupant protection in large school buses: strong, well-padded, well-anchored, high-backed, evenly spaced seats. (The compartmentalization standard does not apply to small school buses, those with a gross vehicle weight rating of less than 10,000 pounds. A number of other safety requirements apply to such buses, including the installation of lap belts.)

Studies by both the National Transportation Safety Board--NTSB (1987) and the Transportation Research Board--TRB (1989) supported the effectiveness of compartmentalization for occupant protection on large school buses. Moreover, these reports recommended that school departments spend their limited resources on other safety improvements, such as pedestrian safety and higher seat backs, rather than on seat belts.³ The debate continues, however.

As mentioned previously, a majority of children killed in school bus-related crashes were outside the bus. Most of these children were struck by the school bus they were exiting or boarding. To reduce the risk to children in the "danger zone" outside the school bus, TRB's 1989 report called for such measures as improved driver training and stop signal arms (now required by NHTSA on all buses). Educating children about safety is also key.

School Children Riding Transit Buses

On November 26, 1996, near Cosmopolis, Washington, a utility truck fatally injured a 10-year-old child who darted from behind a transit bus transporting him home from school. During its investigation of the crash, NTSB determined that school children who ride in transit buses are not provided the same level of safety as those who use school buses, because of differences in the buses' operational practices and equipment (NTSB 1997). For example, while bus drivers are responsible for the safety of children on school buses, even when they are boarding and exiting, children riding transit buses are responsible for their own safety. Moreover, transit buses do not have the identifying markings and flashing lights used on school buses to signal motorists that the bus is carrying children, nor are motorists required to stop for transit buses unloading passengers.

Estimates of the number of children using transit to get to and from school vary. The American Public Transit Association estimates that about 2 million children used transit buses for at least some of their school-related trips in 1994, or approximately 8 percent of public school students. The NPTS indicates that just 2 percent of person trips for school involve transit.

Vans in School Transportation

Some school districts purchase or lease passenger vans to transport students. If these vans carry more than 10 passengers, they are subject to federal safety standards for school buses, and need modifications to make them suitable for school bus use. Vans that have not been modified to conform to the federal safety standards are considered "noncomforming" vans. Federal law prohibits dealers from selling or leasing new nonconforming vans for transporting students.

Passenger van manufacturers have notified their dealers of the federal law against selling or leasing nonconforming vans to schools for pupil transportation (NASDPTS n.d.). Dealers are responsible for determining the intended use of the van and are subject to substantial penalties for violations. Federal regulations, however, apply only to the manufacture and sale or lease of new vehicles. Moreover, although there are no statistics on the number of school children who ride in such vans, a 1997 survey by the National School Transportation Association found that 22 states allow the use of 11- to 15-passenger vans for pupil transportation (NSTA 1999). Just two of these states require that the vans meet the minimum school bus safety standards.

PEDESTRIAN AND BICYCLE SAFETY

Walking and bicycling are a means of transportation used by many people for recreation and exercise. In 1997, there were more than 6,100 pedestrians and bicyclists killed in crashes with motor vehicles. Although pedestrian and bicycle fatalities have declined more rapidly than motorist fatalities since 1976 (see figure 4-4), this does not necessarily mean that the risk has decreased. The NPTS shows that walking trips declined from 9.3 percent to 5.5 percent of trips between 1977 and 1995, while bicycle trips increased from 0.6 percent to 0.9 percent during that period (Pickrell and Schimek 1997).

Although pedestrians and bicyclists are frequently combined in the category of nonmotorized transportation, they are very different in terms of demographics, sources of injuries, and legal responsibilities. Bicyclists travel at three to five times the speed of pedestrians, and are vehicle operators both in terms of physics and according to traffic laws. For example, bicyclists are expected to use the roadway in the direction of traffic. Pedestrians, on the other hand, are expected to use the sidewalk, or, if none, walk facing traffic.

Pedestrian Safety

Pedestrians are particularly vulnerable in a collision with motor vehicles. This is one reason that walkers accounted for only 77,000 (2 percent) of the 3.3 million highway injuries in 1997, but 5,321 (13 percent) of the 42,013 highway fatalities (USDOT NHTSA 1999). Six percent of pedestrians involved in injury crashes were killed and 24 percent had incapacitating injuries; by comparison, only 1 percent of motor vehicle occupants involved in injury crashes were killed and 12 percent received incapacitating injuries (calculated from USDOT NHTSA 1998b, table 53).

Some types of pedestrians are disproportionately represented in the casualty statistics, including children, the elderly, people walking after dark, and intoxicated pedestrians. The relative risks of these groups are unknown, since little information is available about their exposure, that is, the extent and circumstances of their walking. Children 5 to 15 years old accounted for 16 percent of the U.S. population in 1997 but 29 percent of pedestrian injuries in 1997 and 9 percent of pedestrian fatalities.

The large number of children injured may be due to a combination of frequent walking and actions that make them vulnerable to collisions. Older people may have lower than average exposure to traffic dangers as pedestrians, but if they are involved in a crash they have a high rate of mortality. Adults 65 years and older accounted for 13 percent of the population, 8 percent of injuries, but 22 percent of fatalities (USDOT NHTSA 1998b, table 95).

It is estimated that 29 percent of pedestrians killed in crashes in 1997 were intoxicated (with a blood alcohol concentration of 0.10 or more). The rate was highest for pedestrians 25 to 34 years old (50 percent were intoxicated) and pedestrians aged 35 to 44 (47 percent) (USDOT NHTSA 1998e). Alcohol use by motorists is also a factor related to fatal pedestrian-motor vehicle collisions. NHTSA estimates that 12 percent of drivers in such collisions were intoxicated. The percentage of intoxicated pedestrians 14 years old and over killed by motor vehicles decreased from 39 percent in 1982 to 32 percent in 1997 (USDOT NHTSA 1998e, table 20).

Visibility is a major factor in pedestrian fatalities. In 1997, two-thirds of pedestrian fatalities occurred during low-light conditions (dusk, dawn, or dark). Among pedestrians 21 to 44 years old, 81 percent of fatalities happened in low-light conditions (calculated from USDOT NHTSA 1997).

The major circumstances of motor vehicle-pedestrian collisions are shown in table 4-4. Most intersection collisions involve pedestrian dash-outs, turning vehicle conflicts, and motorist traffic signal control violations. Midblock crossings are the next most common type. About half of these incidents involved pedestrians darting out suddenly. Almost half of the collisions involving a backing vehicle or a pedestrian not in the roadway occurred in public or private parking lots (Hunter et al. 1996).

Based on the traffic rules and the descriptions in police accident reports, researchers judged that in car-pedestrian collisions the pedestrian was solely at fault 43 percent of the time, the motorist solely at fault in 35 percent of collisions, and both were at fault 13 percent of the time (calculated from Hunter et al. 1996; in the remaining crashes, the culpability of one or both parties was unknown). Many of the pedestrians at fault on the basis of traffic rules were children. For pedestrians 15 years and older, only 33 percent were judged to be at fault (Hunter et al. 1996, 26).

Bicycle Safety

Bicycling accounts for many fewer deaths than walking, but nearly as many traffic injuries. In 1997, 814 bicyclists were killed and 58,000 were injured in traffic crashes (USDOT NHTSA 1999). Although 90 percent of bicycle fatalities involve a collision with a motor vehicle, most bicycle injuries do not involve a motor vehicle. There are 500,000 bicycle-related emergency room visits annually (Tinsworth et al. 1993). The most common types of bicycle mishaps leading to injury were falls and collisions with fixed objects. Collisions with motor vehicles accounted for about 15 percent of emergency room visits for bicycle injuries (Rivara et al. 1996). Since states do not generally record data about crashes (on-or offhighway) that do not involve motor vehicles, most bicycle crashes are definitionally excluded from traffic injury data.

As with pedestrian data, exposure data for bicycling is limited, which makes it difficult to make statements about changes in risk over time or for different behaviors. For example, it is notable that fewer than one-third (31 percent) of the bicyclists killed in traffic crashes in 1997 were 15 years old or younger, a large change compared with the 1975 figure of 68 percent (see figure 4-5). This change may be the result of an increase in bicycling by adults, a decrease by children, or a change in their relative safety rates. Either the motorist or bicyclist was considered to be intoxicated in 25 percent of the traffic crashes that resulted in bicyclist fatalities in 1997. The bicyclist was intoxicated in 17 percent of all such bicyclist fatalities (USDOT NHTSA 1998d).

A special analysis of police accident reports from six states found the bicyclist alone to be at fault in 50 percent of the crashes, the motorist alone in 28 percent, and both in 14 percent. There was insufficient information to determine fault in the remaining cases (Hunter et al. 1996, table 39). Table 4-5 shows the most common circumstances of car-bicycle collisions. Most crashes occurred when the bicyclist and motorist were on intersecting paths or when one was turning. Only 9 percent involved motorists overtaking cyclists, and 36 percent of these occurred under low-light conditions when visibility of the cyclist may have been a factor (Hunter et al. 1997).

Nearly one-third (32 percent) of bicyclists involved in crashes were riding against traffic (Hunter et al. 1996, table 37). Since most cyclists do not ride against traffic, this figure suggests that this behavior is very risky. In fact, a study that calculated relative risk based on exposure rates found that bicycling against traffic increased the risk of a collision with a motor vehicle by a factor of 3.6 (Wachtel and Lewiston 1994).

There are a number of steps that can be taken to increase bicycle safety:

• wearing a helmet reduces the risk of head and brain injury among cyclists by about 70 percent (Rivara et al. 1996),
• obeying traffic rules,
• using lights and reflectors at night, and
• establishing more bicycle lanes on or near urban roads.

Riding on sidewalks does not reduce the risk of injury; in fact, the relative risk of injury when riding on sidewalks is much greater than when riding on roads (Aultman-Hall and Hall 1998; Wachtel and Lewiston 1994). The greater danger of sidewalk bicycling can be explained by the presence of pedestrians and other obstructions, the danger from turning vehicles at intersections with roads and driveways, and the inexperience of bicyclists who habitually use sidewalks.

TERRORISM AND TRANSPORTATION SECURITY

The past several years have seen mounting concern over the security of transportation facilities and operations, both in the United States and overseas. This has been stimulated both by criminal activity in transportation-related locations, as well as terrorist activity in several parts of the world.

The United States has not escaped this trend. Americans have witnessed worrisome bombings involving parked vehicles at major public facilities: the World Trade Center in New York City (1993), which killed 6 people; and the Murrah Federal building in Oklahoma City (1995), which killed 168 people and injured a total of 759 people. Overseas, U.S. military barracks in Saudi Arabia (1996) and our diplomatic facilities in Dar es Salaam, Tanzania, and Nairobi, Kenya (1998) were attacked by vehicle bombs, leading to more deaths, injuries, and damage.

Credible terrorist threats against many other overseas U.S. government facilities are continually being identified and evaluated, and transportation-related attacks have occurred in other nations. These include the release of sarin nerve gas in the Tokyo subway system (1998), as well as assaults against transit vehicles and systems in France, Israel, China, and several Latin American nations.

The State Department's Office of the Coordinator for Counter-Terrorism issues an annual report on Patterns of Global Terrorism (USDOS 1999). In recent years, the incidence of international terrorism has fluctuated greatly. Although the number of incidents has remained relatively constant at about 300 to 400 annually, the casualties resulting from these events have varied widely from under 1,000 to nearly 7,000 (see table 4-6). A single incident can range from the attempted kidnapping of an individual to a massive bombing of a public facility with numerous casualties and significant destruction.

The most recent report lists 273 acts of international terrorism in 1998, a decrease of 31 from 1997. However, these attacks resulted in the largest number of casualties-741 deaths and 5,952 injuries for a total of 6,693-in any year on record. More than 5,500 casualties were related to the bombings of the U.S. embassies in Nairobi and Dar es Salaam. The 12 U.S. citizens who died in the Nairobi incident were the only Americans to lose their lives to terrorism in 1998. About 40 percent of the 1998 incidents-or 111 of the 273 events-involved U.S. targets. Most of these were bombings of U.S. oil company pipelines in Colombia.

According to DOT's Federal Aviation Administration, there were 216 criminal incidents involving attacks against civil aviation around the world from 1993 through 1997. About 40 percent of the incidents (87) involved hijackings. Another 50 involved attacks at airports, and 32 involved attacks at off-airport facilities (USDOT FAA 1997).

Several recent initiatives by Congress and the Executive Branch emphasized security issues. In 1997, the Commission on Aviation Safety and Security recommended 77 improvements in aviation safety, security, disaster response, and air traffic control (USEOP 1997). Agencies responsible for implementing these recommendations are the Departments of Defense, State, and Transportation; the Federal Bureau of Investigation; the National Transportation Safety Board; the U.S. Customs and Postal Services; and the Bureau of Alcohol, Tobacco and Firearms.⁴

In July 1996, Executive Order 13010 established the President's Commission on Critical Infrastructure Protection (PCCIP). PCCIP conducted a comprehensive review of both the physical and electronic, or "cyber," vulnerabilities of five of the nation's critical infrastructure sectors, defined as systems that, if damaged or disrupted, would have serious negative consequences to economic activities or national security. These five sectors were: information and communications; banking and finance; energy, including electrical power, oil, and gas; vital human services, such as emergency medical response; and physical distribution, including transportation as its major component.

PCCIP's report, issued in October 1997 (USEOP 1997) contained recommendations to improve the nation's response capability. PCCIP identified three areas where transportation was potentially vulnerable. The first would exist if the satellite-based Global Positioning System became the sole source for radionavigation for aircraft landings by the year 2010, as scheduled. The second is related to vulnerabilities as a result of the modernization of the National Airspace System. The third concern arises from the potential vulnerability of advanced electronic information and communications systems used throughout the transportation system. For example, disruptions in computer-based systems used in pipeline operations to monitor flow and pressure and issue automatic electronic commands to remote pumping stations and valves could lead to pipeline failures, environmental damage, and supply shortages. In May 1998, Presidential Decision Directive (PDD) 62 on Combating Terrorism (USEOP 1998a) and PDD 63 on Protecting America's Critical Infrastructure (USEOP 1998b) established a number of new interagency organizations to oversee the implementation of the PCCIP recommendations.

INTERNATIONAL TRENDS IN ROAD SAFETY

Long the dominant mode of travel in the United States, motor vehicles are becoming a preferred means of transportation both in other industrialized countries and in the developing world. In Western Europe, despite its extensive and well-traveled transit and rail networks, automobiles account for more than 86 percent of passenger-kilometers traveled (ECMT 1998).

In 1996, more than 671 million vehicles were registered in the world, up from 411 million in 1980. Although the United States has the most vehicles by far, well over 200 million, its share of the world total declined from 44 percent in 1970 to 31 percent in 1995 (AAMA 1994; 1998). Asfigure 4-6 shows, in 1960 there were 100 million passenger cars registered in the world; by 1995, there were 480 million. Proportionally, the growth in commercial vehicles was even greater, growing from 30 million in 1960 to 170 million in 1995.

World Motor Vehicle Fatalities

Data on the number of fatalities and injuries involving motor vehicle crashes worldwide are limited. Britain's Transport and Road Research Laboratory (TRRL) estimates that 300,000 persons die and 10 million to 15 million persons are injured each year in road crashes around the world (TRRL 1991). Other estimates are higher. One estimate, cited by the International Federation of Red Cross and Red Crescent Societies, is that 500,000 people died in traffic crashes in 1990, and 15 million were injured worldwide (IFRC 1998, 20). According to the World Health Organization (WHO), road traffic accidents in 1998 were the largest cause of ill health or early death for males between the ages of 15 and 44 worldwide, and the second largest cause in developing countries. WHO warned that the burden from traffic accidents would likely increase, especially in developing countries (WHO 1999, 17-18).

Although some countries have made great strides in recent years to reduce vehicle-related deaths and injuries, in others, the number of vehicles is growing far faster than the physical, legal, and institutional infrastructure needed to safely accommodate them. Table 4-7 shows the number of motor vehicle fatalities in selected countries in 1980 and 1996. Most of the countries cited in the table had fewer fatalities in 1996 than 16 years earlier. Many of these countries improved motor vehicle safety through better road design, planning, and operations. Conversely, countries with a greater number of deaths in 1996 are generally those in which vehicle use has risen both quickly and relatively recently.

Safety Issues in Developing Countries

While road safety in most industrialized countries is improving, for many newly developed or developing nations safety problems are worsening. For example, between 1968 and 1985, the number of motor vehicle fatalities in some African countries rose by more than 300 percent and in Asian countries by over 170 percent. In that same period, the number of motor vehicle fatalities in industrialized countries decreased by 25 percent (TRRL 1991).

According to Britain's TRRL, growth in motorization and urbanization in emerging and developing nations has increased traffic on roads that were never designed to carry the volumes that they do today. Moreover, unplanned growth has led to incompatible land use in sprawling urban areas and created significant driving hazards. In many countries, poor road conditions, badly designed intersections, and inadequate protection have exacerbated these dangers for pedestrians (TRRL 1991).

TRRL identifies at least four conditions contributing to road safety problems in developing countries:

• Inadequate design standards. In many developing countries, highway design standards are either outdated (sometimes going back to colonial times) or inappropriate (usually because standards in industrialized countries are applied without considering local needs). Such standards may ignore pedestrians or other nonmotorized transportation or may be too costly for countries to afford.
• Limited resources. Developing nations often lack the engineers and other professionals, as well as the financial resources, needed to modify design standards to accommodate local conditions or to properly maintain roads and other infrastructure.
• Operational and control deficiencies. Operational practices have not kept pace with road building in the developing world. For example, roads are often poorly maintained, traffic signage may be inadequate, walkways for pedestrians are often nonexistent or in poor condition, and control measures to channel vehicles are rarely available (TRRL 1991, 4).
• Inadequate driver training. Drivers in developing countries typically lack the experience of their counterparts in the industrialized world. Many such drivers have never been adequately tested or trained. Moreover, enforcement of traffic laws in developing countries is often ineffective and driver compliance poor.

Safety improvements implemented in the industrialized world have the potential to significantly reduce road hazards in less developed nations. According to TRRL, among the most successful strategies are more safety-conscious road planning and elimination of hazardous locations on roadways through traffic engineering.

DATA NEEDS

While a great deal is known about transportation accidents and incidents and the resulting fatalities, injuries, and property damage, there is much that is not known. This section briefly describes areas where the data available at present are insufficient.

Safety data are, broadly speaking, of two types: outcome data, such as numbers of accidents, fatalities, and injuries; and exposure data, typically expressed as vehicle-miles, passenger-miles, or aircraft departures. The latter are traditionally harder to collect. The following is a list of major exposure data needs:

• exposure measures for the commercial waterborne mode and for pipelines;
• a better exposure measure for recreational boating and better estimates of the number of registered boats;
• data separated by air taxi and general aviation exposures (number of landings), as these data are available only in a combined form since 1994;
• because the common exposure measure for the highway mode, vehicle-miles traveled (vmt), is derived from a sample, error bars on vmt would be very useful in interpreting fatality and crash rates; and
• measures of exposure of specific populations (e.g., elderly drivers or children) and exposure to specific conditions (e.g., adverse weather) are not available, particularly for the highway mode.

Some outcome data are also either not available, or available in combination with fatalities, injuries, and incidents, which are not strictly transportation-related. Some of these data needs are:

• transit and rail fatalities presented in a clear, easy-to-understand format, distinguishing occupants from nonoccupants, and transportation incidents (e.g., collisions and derailments) from nontransportation incidents (e.g., homicides).
• property damage from highway crashes.

REFERENCES

American Automobile Manufacturers Association (AAMA). 1994. World Motor Vehicle Data. Washington, DC.

_____. 1998. Motor Vehicle Facts & Figures, 1998. Washington, DC.

Aultman-Hall, L. and F.L. Hall. 1998. Ottawa-Carleton Commuter Cyclist On- and Off-Road Incident Rates. Accident Analysis and Prevention 30, no. 1:29-43.

European Conference of Ministers of Transport (ECMT). 1998. Trends in the Transport Sector 1970-1996. Paris, France: OECD Publications Service.

Hunter, W.W., J.C. Stutts, and W.E. Pein. 1997. Bicycle Crash Types: A 1990's Informational Guide, FHWA-RD-96-104. Washington, DC: U.S. Department of Transportation, Federal Highway Administration.

Hunter, W.W., J.C. Stutts, W.E. Pein, and C.L. Cox. 1996. Pedestrian and Bicycle Crash Types of the Early 1990s. Washington, DC: U.S. Department of Transportation, Federal Highway Administration.

International Federation of Red Cross and Red Crescent Societies (IFRC). 1998. World Disasters Report 1998. Geneva, Switzerland.

Joksch, H. 1998. Fatality Risks in Collisions Between Cars and Light Trucks, DOT/HS-808-802, prepared for the U.S. Department of Transportation, National Highway Traffic Safety Administration. University of Michigan Transportation Research Institute. October.

National Association of State Directors of Pupil Transportation Services (NASDPTS). n.d. Passenger Vans Used as School Buses, position paper. Available at www.stnonline.com/ core2_files/bbs/nasdpts/sd_pp1.htm.

National School Transportation Association (NSTA). 1999. NSTA Transportation Survey. Available at www.stnonline.com/core2_files/statistics/vans11ormore.htm.

National Transportation Safety Board (NTSB). 1987. Crashworthiness of Large Poststandard School Buses, Report No. NTSB/SS-87/01. Available at www.stnonline.com/core2_files/ bbs/ntsb/largebusstudy.htm.

_____. 1997. Highway Accident/Incident Summary Report: Collision with a Pedestrian by a Utility Truck Near Cosmopolis, Washington, November 26, 1996, Report No. NTSB/ HAR-97/01/SUM. Washington, DC.

Pickrell, D. and P. Schimek. 1997. Trends in Personal Motor Vehicle Ownership and Use: Evidence from the Nationwide Personal Transportation Survey. Washington, DC: U.S. Department of Transportation, Federal Highway Administration. Also available at www-cta.ornl.gov/npts/1995/Doc/publications.shtml.

Rivara, F.P., D.C. Thompson, and R.S. Thompson. 1996. Circumstances and Severity of Bicycle Injuries. Seattle, WA: Snell Memorial Foundation, Harborview Injury Prevention and Research Center.

Tinsworth, D., C. Polen, and S. Cassidy. 1993. Bicycle-Related Injuries: Injury, Hazard, and Risk Patterns, Technical Report. Washington, DC: U.S. Consumer Product Safety Commission.

Transportation Research Board (TRB), National Research Council. 1989. Improving School Bus Safety, Special Report 222. Washington, DC.

Transport and Road Research Laboratory (TRRL), Overseas Development Administration. 1991. Towards Safer Roads in Developing Countries: A Guide for Planners and Engineers. United Kingdom: Ross Silcock Partnership.

U.S. Department of State (USDOS). 1999. Patterns of Global Terrorism: 1998, Publication 10610. Washington, DC. April.

U.S. Department of Transportation (USDOT). 1998. White House Commission on Aviation Safety and Security: The DOT Status Report. Available at www.dot.gov/affairs/whcoasas.htm, as of 7 June 1999.

U.S. Department of Transportation (USDOT), Federal Aviation Administration (FAA). 1997. Criminal Acts Against Civil Aviation. Available at http://cas.faa.gov/cas/crim97.htm.

U.S. Department of Transportation (USDOT), National Highway Traffic Safety Administration (NHTSA). 1997. Fatality Analysis Reporting System. Available at www.fars. nhtsa.dot.gov.

_____. 1998a. School Bus Safety: Safe Passage for America's Children. Available at www.nhtsa.dot.gov/people/injury/buses/schbus/schbussafe.html.

_____. 1998b. Traffic Safety Facts 1997. Available at www.nhtsa.dot.gov/people/ncsa/ TSF97.html.

_____. 1998c. Traffic Safety Facts 1997: Large Trucks. Available at www.nhtsa.dot.gov/people/ncsa/Truck97.html.

_____. 1998d. Traffic Safety Facts 1997: Pedalcyclists. Available at www.nhtsa.dot.gov/people/ncsa/PedCycle97.html.

_____. 1998e. Traffic Safety Facts 1997: Pedestrians. Available at www.nhtsa.dot.gov/people/ncsa/Ped97.html.

_____. 1998f. Traffic Safety Facts 1997: School Buses. Available at www.nhtsa.dot.gov/people/ncsa/factshet.html.

_____. 1999. Traffic Safety Facts: Overview 1998. Available at www.nhtsa.dot.gov/people/ncsa/factsheet.html

U.S. Executive Office of the President (USEOP). 1997. Final Report of the White House Commission on Aviation Safety and Security. Washington, DC. February.

_____. 1998a. Presidential Decision Directive 63: Combating Terrorism, Fact Sheet. Washington, DC. May.

_____. 1998b. Presidential Decision Directive 62: Protecting America's Critical Infrastructure, Fact Sheet. Washington, DC. May.

Wachtel, A. and D. Lewiston. 1994. Risk Factors for Bicycle-Motor Vehicle Collisions at Intersections. ITE Journal. September.

World Health Organization (WHO). 1999. World Health Report 1999. Geneva, Switzerland.

FOOTNOTES

¹ For a more detailed discussion of mode-by-mode trends, see chapter 3 in the past three issues of the Transportation Statistics Annual Report (TSAR). Cross-modal topics discussed in these issues include: in TSAR98--transportation and the elderly, causes of crashes, and alcohol and drug involvement; in TSAR97—child safety, safety of transportation workers, and hazardous materials; in TSAR 96--fatigue, highway-rail grade crossings, safety technologies, and alcohol and drug involvement.

² The NPTS collected travel information from a sample of U.S. households for calendar year 1995 and then used demographic statistics to expand the results to nationwide estimates. Among other types of data, the NPTS collected information on trips of a daily nature (e.g., commuting, shopping, and transporting children to school), sorted by the type of vehicle, the nature of the trip, trip length, and other parameters. The NPTS thus allows users to calculate both the number of person trips and the number of person-miles traveled by almost any subgroup of the population for almost any type of trip.

³ TRB, however, concluded that properly used seat belts on large buses built after 1977 may reduce school bus passenger deaths or injuries in crashes by up to 20 percent (TRB 1989, 83).

⁴ DOT’s annual status report (USDOT 1998) on the implementation progress is available on the Internet at www.dot.gov/affairs/whcoasas.htm.