The injuries and fatalities associated with school bus crashes are some of the most emotional and tragic types of events that occur. They also often receive a large amount of publicity. The background and rationale behind the current regulatory requirements (high seat backs and compartmentalization) that provide occupant protection in large school buses are often not well understood by the public. Some believe that seat belts would have a positive effect on school bus safety, while others believe the opposite. For example, proponents for seat belts in large school buses raise a number of specific issues regarding the performance of compartmentalization in side impact and rollover crashes, the potential benefits of school bus seat belts, and the level of research that evaluated the effects of seat belts in school buses. This paper presents 1) data on school bus fatalities and injuries, and 2) a summary of the research relative to the effects of seat belts in school buses. The issues regarding compartmentalization, lap belt effectiveness, carryover effects of seat belt use in school buses, and improved student behavior due to seat belt use are also discussed.

1) School Bus Fatalities and Injuries

The National Highway Traffic Safety Administration (NHTSA) uses the Fatality Analysis Reporting System (FARS) to obtain information on the number of fatalities in school buses each year and reports the results in a publication - Traffic Safety Facts - School Buses. FARS contains data on all fatal traffic crashes within the 50 states, the District of Columbia, and Puerto Rico. The data system was conceived, designed, and developed by the agency's National Center for Statistics and Analysis (NCSA) to assist the traffic safety community in identifying traffic safety problems, developing and implementing vehicle and driver countermeasures, and evaluating motor vehicle safety standards and highway safety initiatives.

NHTSA contracts with an agency in each state to provide information on fatal crashes. FARS analysts are state employees who extract the information and put it in a standard format. Each FARS analyst attends a formal training program, and also receives on-the-job training. Data on fatal motor vehicle traffic crashes are gathered from the states' own source documents.

Analysis of FARS reveals that from 1991 to 1996 there have been 41 fatalities of passengers other than the driver in school bus body type vehicles. Of these 41 fatalities, six fell from the school bus, seven were a result of a train traveling at a high rate of speed and crashing into the school bus, eight occurred from collisions with poles or trees, one resulted when the bus overturned, 18 were a result of collisions with other vehicles, and one is listed as a non-collision. Of the 34 fatalities that resulted from collisions, 13 were from frontal crashes, 11 were from rear collisions, and 10 were from side impacts.

To provide insight into the amount of injuries in school buses, NHTSA uses data from the General Estimates System (GES) which is a nationally representative sample of police reported motor vehicle crashes of all types of motor vehicle incidents, from minor to fatal. The system began operation in 1988, and was created to identify traffic safety problem areas, provide a basis for regulatory and consumer initiatives, and form the basis for cost and benefit analyses of traffic safety initiatives. The information is used to estimate how many motor vehicle crashes of different kinds take place, and what happens when they occur.

By restricting attention to police-reported crashes, the GES concentrates on those crashes of greatest concern to the highway safety community and the gener al public. These crash reports are chosen from 60 areas that reflect the geography, roadway mileage, population, and traffic density of the U.S. The data are collected by making weekly visits to approximately 400 police jurisdictions in areas across the United States, where they randomly sample about 50,000 police reports each year.

GES estimates from 1988 to 1996 indicate that there are approximately 8,500 in juries in school buses each year. The overwhelming majority of these injuries, approximately 7,285, would be considered minor. There are approximately 350 injuries annually that would be considered serious. These numbers do not fluctuate widely from year to year, as some suggest. Statistical methods require that the sample size be as large as possible to reduce the amount of error in the estimate. These data are the most reliable available crash information. They are based on proven statistical methods that provide the most accurate a nalysis of injuries to occupants in school buses.

2) Research to Evaluate the Effects of Seat Belts in School Buses

Currently, the agency is working closely with the National Transportation Safety Board (NTSB) on a computer simulation of the Monticello, Minnesota, school bus crash event. This simulation should help to provide insight into the occupant kinematics during the crash. From this information it may be possible to ascertain whether the use of lap belts would have been beneficial at reducing injuries or fatalities in that crash.

In Fiscal Year 1998, Congress appropriated 700,000 to evaluate alternative safety restraint bar devices on school buses. NHTSA has recently received approval from both the House and Senate Appropriations Committees to use this money to evaluate other occupant restraint devices in school buses, such as lap belt. On June 9, 1998, President Clinton signed the landmark Transportation Equity Act for the 21st Century (TEA21). Provisions in TEA21 call for the agency to spend up to 200,000 on school bus occupant safety research. Other provisions include conducting a study concerning the safety of all transportation modes used in school transportation, i.e., school buses, transit buses and other passenger vehicles. In NHTSA's continuing efforts to make school buses even safer, a bus safety research plan is currently being developed to most efficiently utilize these resources. This plan will look at available information and develop concepts to improve the occupant protection provided by school buses. Testing will be performed to evaluate alternative occupant protection conce pts and to support any regulatory changes that may come about as a result of this research.

The following is a review of previously conducted school bus safety research:

School Bus Passenger Protection; Society of Automotive Engineers, Institute of Transportation and Traffic Engineering, University of California at Los Angeles, by Derwyn M. Severy, Harrison M. Brink and Jack Baird, 1967.

In 1967, the University of California at Los Angeles (UCLA), Institute of Transportation and Traffic Engineering, published a report, "School Bus Passenger Protection," on crash testing it performed on school buses. These crash tests are often cited during discussions of whether lap belts are necessary for occupant protection in large school buses. Three crash tests were performed: a head-on collision between two school buses, each traveling 30 mph; a stationary rear impacted school bus by a passenger car traveling at 60 mph; and a stationary school bus side impacted on its right side by a passenger car traveling at 60 mph. The researchers tested a variety of restraint options, lap/shoulder belts, lap belts, restraint bars, and air bags with a variety of dummy types and sizes representing adult, 3, 6 and 13-year-old people. These crash tests used two prestandard (manufactured before April 1977) large school buses: a 1944 Mack-Superior Coach and a 1965 GMC-Superior Coach. Ten different seat types were tested; conventional seat found in a 1965 Superior school bus, a fiberglass molded seat, an automotive type bucket seat with headrest, and an airline seat.

In 1972, a second set of school bus crashes was conducted at UCLA (Wojcik and Sande). A 1969, 60-passenger Superior school bus was crashed head-on into a 2-ton dump truck, each traveling at 30 mph. A side impact test was also performed on the same stationary school bus with a 1967 Ford sedan traveling at 60 mph. Again, the passenger size, restraint type and seat design were manipulated.

The crash performance and interior design of prestandard school buses are not comparable to that of poststandard (manufactured April 1977 or later). None of the combinations of seating types were representative of that used in modern school buses.

Development of a Unitized School Bus, Volume 1 - Summary Report; DOT-HS-802-004, National Highway Traffic Safety Administration, U.S. Department of Transportation, Washington, D.C., by L. Adams, et al., 1976.

This report summarizes the development of design concepts used for passively providing occupant protection for school bus passengers. Analysis and develop ment tests indicated the feasibility of passively providing uniform levels of protection to seated occupants, ranging in size from a six-year-old child to a 50th percentile adult male in frontal and rear 30 mph impacts. The report also summarizes the results of a 30 mph side impact collision of a school bus body and a rigid pole at 30 mph. Another of the design goals of this study was a unitized bus configuration where the body and chassis frame are integrated in a single structure, in contrast with the typical school bus configuration where the body is bolted onto a chassis frame.

This research serves as the basis for the current occupant protection require ments in FMVSS No. 222, School Bus Seating and Crash Protection. This research demonstrated that in side impacts with a rigid pole, uniform occupant protection could not be provided.

School Bus Passenger Seat and Lap Belt Sled Tests; DOT HS-804-985, National Highway Traffic Safety Administration, U.S. Department of Transportation, Washington, D.C., by A. R. Bayer, 1978.

In 1978, NHTSA reported on a series of sled tests it conducted to obtain additional data in support of rulemaking activities related to FMVSS No. 222. In response to comments received during rulemaking, a sled test program was conducted to evaluate the performance of various production regular route school bus seats and activity school bus seats. The testing variables consisted of seat spacing, test speed, differing dummy sizes, and the use/non-use of lap belts. Sled tests were performed at 15 and 20 mph.

Again, the crash performance and interior design of prestandard school buses were not comparable to that of poststandard (manufactured April 1977 or later).

School Bus Safety Study - Volume I; Traffic Safety Standards and Research, Crashworthiness Section, Transport Canada, Quebec City, Quebec, Canada, by G. N. Farr, 1985.

In 1985, Transport Canada published the result of three school bus crash tests. All three school buses were subjected to a 30 mph barrier test. The test vehicles were a large conventional school bus (GVWR > 10,000 lb), a 66-passenger Blue Bird, and two small school buses (GVWR < 10,000 lb), a 22-passenger Thomas Minotour on a Ford chassis and a 20-passenger Campwagon conversion from a B350 Dodge van. Each bus contained six 5th percentile adult female dummies. Three of the dummies were restrained with type I lap belts and the other three were unrestrained. The conclusions of this study are as follows:

• The test results showed the potential for head injury in frontal collisions increased when lap belts were used. Careful deliberation must be exercise d before deciding whether or not to add lap belts to existing designs of occupant protection systems found on today's school buses.

• For the large school bus all dummies exhibited HIC values of less than 1,000. The HIC for the belted dummies were three times greater than the HIC for the unbelted dummies.

• For the small school buses, all HIC values for the belted dummies exceeded 1,000 and all HIC values for the unrestrained dummies were less than 1,000.

• All lap belted dummies experienced severe rearward neck flexure. This in itself was judged to cause at least serious injury.

In 1987, NTSB published a report on 43 serious school bus crashes that it inv estigated to determine the effect of seat belts. Those crashes included frontal, side and a large number of rollover crashes. For each crash, an evaluation was made to determine whether the use of seat belts would have made a difference in the injury levels of the school bus occupants. NTSB concluded that:

• School bus occupant deaths and serious injuries sustained by survivors for the most part were attributable to the occupants' seating position being in direct line with the crash forces.

• It is unlikely that the availability of any type of restraint would have improved the injury outcome.

• Rollover incidents in the NTSB study were associated with higher levels of school bus passenger injuries than non-rollovers but to a much smaller degree than anticipated. Nearly 86 percent of all the school bus passengers were either uninjured or received only minor injuries.

• The slight increase in school bus passenger injury severity associated with rollover incidents was due primarily to school bus rollovers preceded by collisions. The initial impact, not the rollover, was responsible for the higher injury levels.

• Lap belt use probably would have made no change in the total number of fatalities to school bus passengers ... possibly one more fatality would have resulted.

• At best, lap belt use probably would have reduced somewhat the injuries of 8 of the 24 school bus passengers with serious injuries and made no difference to the majority.

• At worse, lap belt use would have increased the injuries to almost as many passengers with serious injuries as it improved.

• Lap belt use probably would have worsened the outcome of one-fifth [20 percent] of the 58 school bus passengers with moderate injuries.

Special Report 222: Improving School Bus ; Transportation Research Board, National Research Council, Washington, D.C., 1989.

In 1987 the Surface Transportation and Uniform Assistance Act had a provision that required the National Academy of Sciences (NAS) investigate the principal causes and injuries to school children riding in school buses, the use of seat belts in school buses, other measures that may improve the safety of school transportation and to determine those safety measures that are most effective at protecting school children while boarding, leaving, and riding in school buses. To conduct this study, the National Research Council, the operating agency of the National Academy of Sciences and Engineering, assigned the study to the Transportation Research Board (TRB).

Special Report 222's review of the effectiveness of lap belts at reducing injury in school bus crashes was based on four types of studies: (a) inferences drawn from statistical evaluations of the effectiveness of lap belts in the rear seats of passenger cars, (b) crash tests (UCLA and Canada), (c) sled tests, and (d) analysis of real world crashes. The conclusions were as follows:

• The effectiveness of lap belts in passenger cars suggest that lap belts may reduce the number of fatalities by 20 percent and the number of serious and fatal injuries by 40 percent.

• The school bus crash tests and sled tests reviewed by the committee did not suggest that seat belts (lap type) would or would not be effective in frontal collisions.

• The dummies restrained by lap belts in both tests sustained lower chest loading but higher head-injury-criterion measurements.

• NTSB's in-depth analysis of the effectiveness in severe school bus crashes suggest that little if any benefit is attributable to seat belt use.

• Seat belts, when properly used on large, post-1977 buses, are not inherently harmful and may reduce the likelihood of death or injury to passengers involved in school bus crashes by 20 percent.

• The overall benefits of requiring seat belts on large school buses are insufficient to justify a Federal requirement for mandatory installation.

  The funds used to purchase and maintain seat belts might better be spent on other school bus safety programs and devices that could save more lives and reduce more injuries.

• The following are four issues that are most often raised by proponents for seat belts in large school buses. Discussions of each of these issues based on

i) Compartmentalization of School Children in School Buses Does Not Provide Protection in Side Impact and Rollover Crashes.

Analysis of FARS reveals that from 1991 to 1996 there have been 41 fatalities of passengers other than the driver in school bus body type vehicles. Of these 41 fatalities, one resulted when the bus overturned and 10 were from side impacts.

It cannot be determined if seat belts would have made a positive difference in the injury/fatality outcome merely by looking at the aggregate data. To determine if seat belts would have been beneficial in these crashes, the circumstances of each crash need to be investigated individually. In 1987, NTSB report ed on 43 serious school bus crashes that it investigated and concluded that, for the most part, occupant deaths and serious injuries sustained by survivors were attributable to the occupants' seating position being in direct line with the crash forces. NTSB concluded that the slight increase in school bus passenger injury severity associated with rollover incidents was due primarily to school bus rollovers preceded by collisions. The initial impact, not the rollover, was responsible for the higher injury levels.

ii) The Transportation Research Board Special Report 222 Concluded That a 50 Percent Lap Belt Usage Rate Would Have Reduced Deaths and Injuries by 20 Percent.

In August 1994, the Center for Urban Transportation Research, College of Engineering, University of South Florida, published a report, "To Belt or Not To Be lt, Experiences of School Districts that Operate Large School Buses Equipped with Seatbelts." This study addressed the issue of belt usage, among other things, in school buses. Of the 814 school districts included in the sample, 763 were in New York State primarily because New York began mandating seat belt installation (but not their use) in 1987. They reported that 77.5 percent of the respondents indicated that students use the seat belts 10 percent of the time or less, while only 6.1 percent of the respondents indicated that students use the seat belts 51 percent of the time or more. New Jersey was excluded from the survey because they had only recently mandated seat belts in school buses. NHTSA is not aware of any data that are currently available on New Jersey seat belt use rates.

The TRB study concluded that seat belts may, not would, result in a 20 percent reduction in injuries if a 50 percent usage of the school bus lap belt was obtained. The TRB study did this by inferring that seat belts on large school buses would be effective just because of their effectiveness in passenger cars. Public policy and Federal regulation must be based on data and science. On the other hand, seat belts are effective in small (van-type) school buses because they experience similar crash forces as do passenger cars and other small vehicles. Therefore, NHTSA requires these small school buses to be equipped with seat belts at all designated seating positions, just like cars, and other light trucks and vans.

iii) Children Need to Be Using Seat Belts in School Buses in Order to Reinforce Educational Message and This Will Carry Over to Other Vehicles They Ride In.

In February 1986, NHTSA published a report, "School Bus Safety Belts: Their Use, Carryover Effects and Administrative Issues," DOT HS 806965. The report found no conclusive evidence of a carryover effect between the school bus and the family car. The report concluded:

Parent rules and attitudes, mandatory State car belt use laws, and classroom education programs appeared to play more dominant roles in students' car belt use than bus-belt carryover effects ... As more States enact mandatory car be lt use laws, more students would already use car belts, and hence, fewer will remain to be affected by any possible carry over effects.

The agency is not aware of any additional studies that support any conclusion that the use of seat belts on school buses will carry over to increased seat belt use in other passenger vehicles.

iv) Proper Use of Seat Belts Would Improve Student Behavior on the School Bus, Reduce Driver Distractions, and May Translate into Crashes Avoided.

The Center for Urban Transportation Research study mentioned above concluded that:

Overall, student conduct did not improve while riding in the school buses as a result of the provision of seat belts and, in a few instances, the questionn aire respondents indicated that student conduct actually had become worse. The tabulated data revealed that 90.4 percent of the respondents indicated that student conduct did not improve, while 9.6 percent indicated that it did. In one instance a respondent noted that on-board conduct had improved, but only among elementary school students.

The agency is not aware of any additional studies that support any conclusion that the use of seat belts on school buses will improve student behavior and result in avoiding crashes.