Pedestrian Fatalities and Injuries in the U.S.

Mark D Larsen

July 26, 2010

Recent lobbying efforts by the National Federation of the Blind have convinced the members of the U.S. Senate’s Commerce, Science, and Transportation Committee to include Amendment Sec. 109 in the proposed Motor Safety Act 2010.

In brief, the amendment would “require new electric or hybrid vehicles to provide an alert sound” at under 20 mph. The rationale behind the requirement is that such vehicles are so relatively “quiet” at low speeds that they pose a unique danger to pedestrians, especially the blind who cannot see them approaching.

Oddly, the regulation does not apply to any and all vehicles that might prove too “quiet” for pedestrians to hear, but only electric and hybrid vehicles. To be sure, many automobiles with internal-combustion engines have likewise become increasingly “quiet” during the past several years, thanks to engineering improvements that reduce the already intolerable cacophony of our streets. If the goal of the regulation is to truly render “quiet” vehicles more noticeable by pedestrians, particularly those who are blind, it should identify them by the level of decibels produced at under 20 mph —rather than solely by the drivetrain used. Amendment Sec. 109 fails to do this, and instead targets only electric or hybrid vehicles —no matter how much noise a given model already produces.

One can empathize with the challenges faced by blind pedestrians, and on the surface it seems logical that “silent” vehicles might exacerbate those challenges. However, as drivers who own electric and hybrid vehicles will attest, although they are indeed quieter than cars from past years, they are certainly not “silent.” One has to question, therefore, whether or not the purported danger truly exists or is more the result of mere speculation, conjecture, and unsubstantiated fears.

There have been electric and hybrids vehicles on our streets for many years now. As only one example, the first model of the Toyota Prius hybrid was introduced in the U.S. in 2001, and there are now approximately one million of them on our roads. If electric and hybrid vehicles really do pose a particular danger because of their “quiet” drivetrains, surely NHTSA statistics will show that there has likewise been a proportionate increase in the number of pedestrian fatalities and injuries during those years.

In reality, however, the data from the NHTSA’s Fatality Analysis Reporting System (FARS) do not substantiate that hypothesis. In fact, they reveal that there has actually been a decrease in such incidents, despite the growing number of “quiet” electric and hybrid vehicles on our streets.

Before examining those data, however, it merits mention that readers who reviewed an earlier draft of this summary raised pertinent questions. For example, one suggestion was that it might be helpful to separate fatalities and injuries to determine if both types of accidents show the same trend over time. The FARS obviously focuses on fatalities, so its statistics on mere injuries may or may not be complete. Nonetheless, it is possible to compile such data via the category for “Injury Severity” by purposely unchecking the search variable “Fatal Injury” and checking all the remaining types of injuries. The tables and graphs below now separate those data accordingly.

Another prudent suggestion was to adjust the numbers according to VMT (Vehicle Miles Traveled). That change makes complete sense: the more a vehicle is driven, the greater the risk of an accident. It is worth pointing out, however, that how much a vehicle is driven —and its corresponding accident risk— doesn't necessarily depend exclusively on miles: it could also depend upon time behind the wheel and/or the environment. For example, if one drives 3 hours on rural roads at an average of 75 mph, the distance is 225 miles. Yet if one drives 8 hours in urban traffic, at an average of 20 mph, the distance is only 160 miles. Nonetheless, the risks of an accident are probably much greater in the city over a longer period of time than in the country for a shorter period of time. Regardless, for good or evil, the Federal Highway Administration has decided to use miles traveled to measure “driving,” and thus that is the only option available to adjust the tables and graphs in this summary.

The following Figure 1 plots the VMT gathered from the Federal Highway Administration’s Traffic Volume Trends during the same years as the FARS, i.e., 1994 through 2008. As is evident, VMT have been increasing every year; the sole exception occurred in the last year, 2008 —likely a consequence of the soaring gasoline prices that year.

Below are tables with data from the FARS system, adjusted to the VMT (in billions).


CATEGORY 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 TOTALS
ALL Fatalities 17.26718 17.25836 16.94803 16.41133 15.80990 15.57186 15.26938 15.09156 15.06305 14.83875 14.45209 14.55671 14.16987 13.61683 12.75188 229.076780
ALL Injuries 16.39101 16.67478 16.41781 15.69648 15.05524 14.55207 14.13578 13.88591 13.47706 13.20069 12.61167 12.14587 11.53484 10.71650 9.63997 206.135691
Pedestrian Fatalities 2.32782 2.30458 2.19541 2.07852 1.99162 1.84360 1.73389 1.75286 1.69912 1.65190 1.57726 1.63667 1.59091 1.55083 1.49829 27.433273
Pedestrian Injuries 0.20738 0.22947 0.18775 0.17969 0.16000 0.16536 0.15544 0.15773 0.15972 0.16817 0.12922 0.13617 0.12143 0.11386 0.12423 2.395608
Blind Pedestrian Fatalities 0.00170 0.00083 0.00121 0.00039 0.00190 0.00187 0.00146 0.00072 0.00070 0.00035 0.00067 0.00067 0.00000 0.00000 0.00068 0.013139
Blind Pedestrian Injuries 0.00000 0.00000 0.00000 0.00000 0.00000 0.00037 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.000373

The following graphs help put these statistics into perspective by plotting the data over time. Figure 2 below shows that, although the U.S. population has steadily increased during those 15 years, the total numbers of traffic fatalities and injuries per billion VMT have decreased.

Figure 3 below plots the numbers of pedestrian fatalities and injuries per billion VMT during those same years. It clearly illustrates that, rather than increasing, those types of mishaps have likewise been decreasing for many years. Readers will undoubtedly note that, when a vehicle hits a person rather than another vehicle, the chances of surviving with only injuries are much lower. It is also worth mentioning that pedestrian accidents are relatively few when compared with the total number of accidents: they constitute merely 6.83% of all the fatalities and injuries that occurred during those 15 years.

The data for blind pedestrian fatalities and injuries during those 15 years are so extremely low that they are practically negligible. In fact, to calculate such accidents, Figure 4 uses per trillion VMT, since more often than not they amount to an average of only one or two per year. Like with all pedestrians, the numbers for the blind have likewise been declining over time to the point that in 2006 and 2007 no such fatalities occurred, and the sole injury reported by the FARS occurred in 1999.

The “bottom line” is this: from 1994 through 2008, out of 1,492,637 traffic fatalities and injuries in the U.S., 36 were blind pedestrians. That is less that 0.0025%. Moroever, most of those incidents occurred a decade earlier, before the advent of hybrid vehicles, and not one of them at less than 20 mph.

These statistics beg the question: why are the numbers of pedestrian fatalities and injuries dropping rather than increasing, if the proliferation of “quiet” electric and hybrid vehicles poses such a danger to pedestrians —especially to the blind? The most reasonable explanation is that, through the years, more and more vehicles have improved safety features (elevated third brake lights, front and side airbags, tethers for child seats, better crumple zones, etc.), resulting in fewer fatalities and injuries per billion miles driven. Regardless, these data show no indication that the growing number of hybrid or electric vehicles on our roads during the past decade is causing a proportionate rise in the number of pedestrian accidents —let alone of blind pedestrians. And certainly not at low speed!

Of course, in all fairness, a very remote possibility is that the numbers of pedestrians hit by other types of vehicles are dropping so precipitously that they “mask” a rise in such incidents among hybrid or electric vehicles. The only way to prove that beyond a shadow of a doubt would be for NHTSA to make public all the national data according to drivetrain. So far, they have not done so, but instead have only issued a “selective” report which, by their own admission, is anything but conclusive: it compares pedestrian accidents involving ICE vehicles in general with those involving hybrids in only a few crowded, urban centers —where such mishaps always happen more frequently. See the Noise Off organization's report on this issue.

It is human nature to want to trust and generalize personal anecdote and conjecture more than impersonal numbers. This is especially true if questioning such assertions makes one seem “insensitive” to the needs of the disabled. Nonetheless, unless and until such a “need” is unequivocally substantiated with bona fide objective data, it would be counterproductive for lawmakers to impose Amendment Sec. 109 on the rest of society. It would only add to the noise pollution of our streets, raise the manufacturing costs of electric and hybrid vehicles, and discourage consumers from buying them —all to address a problem which, according to the government’s own statistics, might not exist. We should always have solid empirical evidence to substantiate the need for such a law before enacting it —and not just theory or anecdote.