School transportation is safe, but surveys don’t tell the whole story.

Moments after unloading, a 13-year-old girl is playing near the back of her school bus when she falls and is fatally run over by the rear tires. An 11-year-old girl is struck and killed by a passing vehicle as she tries to cross the road to wait at her bus stop. Four students die when their bus plunges off an Alabama highway overpass.

Should these incidents be included in the school transportation industry’s best safety assessments? Yes, and no.

The first two are absent from the Kansas Department of Education’s survey on loading and unloading fatalities for the 2007-2008 school year. The third — one of the most high-profile crashes in recent history — appears nowhere in the industry’s analysis of the National Highway Traffic Safety Administration Fatal Accident Reporting System (FARS) for the 2006-2007 school year.

These reports indicate the past couple of years were some of the safest years on record, part of an overall trend. According to the Loading and Unloading survey, five students were killed in the school bus “danger zone” last school year, four of whom were struck by passing vehicles. In 2006-2007, seven were killed. The year before, 13. Similarly, an analysis of FARS shows no students killed inside the school bus during the most recent 2006-2007 school year for the first time in a decade.

But neither survey gives a complete picture.

A School Transportation News compilation of national headlines over the last three years shows these fatalities that don’t make it into either report. When combined with the traditional methods, this shines a brighter light on the limited dangers of riding the safest form of transportation.

This year, Texas was the only state not to participate in the voluntary loading and unloading study. But as a result, at least one fatality was not included. According to a local ABC affiliate, an 8-year-old boy was fatally struck by a passing vehicle when he stepped into the roadway in the early morning hours of Sept. 27, 2007. His bus had yet to arrive at the stop.

But a state’s submission does not guarantee accuracy. The incident of the 13-year-old girl run over by the rear tires, as reported by Newsday.com, occurred in New Jersey, a state that participated in the survey. Yet the girl is not included in the fatality figures.

Even if Texas had reported, it’s unclear whether this fatality would have been included because there is no standard definition of a loading or unloading fatality. Traditionally, the survey is seen as addressing only the loading/unloading process. Fatalities that occurred when students were walking to or waiting at their bus stop before the bus had arrived usually go uncounted. The state directors of pupil transportation that submit data ultimately decide which fatalities count.

Nonetheless, according to expert witness Ned Einstein, these bus stop incidents pose a significant danger. STN’s independent analysis of national headlines found at least four incidents where students were killed while walking to or waiting at the bus stop before the bus had arrived that do not appear in the report, despite their state’s participation in the survey.

Other fatalities, like a 16-year-old student who was ejected and died after his bus collided with a dump truck, are omitted by design. The annual loading and unloading survey only examines deaths in the boarding and alighting process, which is thought to be the most dangerous part of the statistically safe process of riding a school bus. All others, including the death of a 7-year-old Worden, Mont., girl who was fatally ejected when her bus ran into a pickup filled with gravel and four Cottonwood, Minn., students who died in the well-known crash between a van and their school bus last February, are excluded.

As an overall assessment of school transportation safety, FARS is no more complete. The most recent 2006-2007 school year figures, prepared by Doug Snyder for the California Association of School Transportation Officials, shows no school bus passenger fatalities. However, in the same year, four students died when their bus careened over the edge of a Huntsville, Ala., freeway overpass.

This, too, may be a case of survey limitations. The analysis looks only at deaths that occur during school transportation hours (between 6 a.m. and 9 a.m. and 2 p.m. and 5 p.m., Monday though Friday). So, the Huntsville incident remains invisible because it occurred on a bus that was carrying students from a regular high school to a vocational school during the school day. Still, the crash was one of the most high-profile incidents in recent years and launched a statewide study of seat belt usage.The STN study is no less flawed. STN uses Google to search some 4,500 national and regional newspapers daily for school transportation-related keywords. This casts a wider net than either survey, but relies on other news sources reporting incidents factually and STN editors to determine whether these are truly school transportation fatalities. For example, a student pedestrian fatality might be flagged in the STN report because the newspaper reported that the student was killed near a school bus stop.

School bus safety has no doubt improved since the survey first loading and unloading started 37 years ago. In the Kansas survey’s first year, 75 students were killed while loading and unloading, far fewer than the number killed this year by any measure. But until the industry develops a standardized method of reporting school bus-related fatalities it will have to rely on this statistical hodgepodge.

Reprinted from the January 2009 issue of School Transportation News magazine. All rights reserved.