Ivars Peterson's MathTrek
The U.S. Department of Transportation's Fatality Analysis Reporting System maintains a database that provides detailed information about 1,284,629 of these fatalities. The information comes largely from police officers called to crash scenes.
Such an enormous database allows researchers to test a wide variety of specific hypotheses about the causes of traffic fatalities. In the current issue of Chance, traffic safety expert Leonard Evans examines the role of sex differences.
Overall, the data show that about 73 percent of all people killed in car accidents are male. For every age, all the way through the mid-90s, male fatalities outnumber female fatalities by a factor of 3 or more. At age 20, for example, there were 14,325 male and 3,238 female drivers killed while traveling alone during the 29-year period.
Interestingly, a similar ratio holds for pedestrian deaths. The period from 1975 to 2003 saw 126,715 male deaths and 55,669 female deaths. Indeed, even among infants and young children, male fatalities invariably outnumbered female fatalities.
"The ratio of male to female pedestrian risk, while greater than one at all ages, increases rapidly to more than three at an age close to puberty," Evans writes. Similar patterns can be seen in the sex and age dependence of arrest data and measured testosterone levels, he contends.
"This paper cannot establish that the effects found are innate," Evans concludes. "However, I believe the simplest and most plausible interpretation is that the effects reported reflect intrinsic behavioral differences between the sexes originating at a hormonal level."
In an accompanying commentary on Evans' paper, statistician Mary Meyer of the University of Georgia raises several concerns about the analysis. She points out that Evans' use of raw numbers is appropriate only if men and women drive approximately the same number of miles under roughly the same conditions.
That isn't the case, Meyer argues. A U.S. Department of Energy study, for example, revealed that men averaged 65 percent more driving miles than women did in 2001. At the same time, while men log more miles, women log more trips, on average. Indeed, data show that more female drivers are involved in minor accidents while more male drivers are involved in severe accidents, Meyer says.
Overall, when driving miles are taken into account, the fatality rate among men is 70 percent (rather than 270 percent) higher, she notes.
"Is the higher fatality rate for miles driven by men an indication of innate sex differences, or due to socialization?" Meyer asks.
Although she doesn't completely rule out a hormone explanation, Meyer suggests that other, more direct social factors, such as alcohol use, help explain the disparity. In 2003, for example, 27 percent of the male drivers in fatal crashes had been drinking, whereas only 12.5 percent of the female drivers had done so. And, of male drivers, 34 percent were wearing no seatbelt versus 24 percent for female drivers.
In another commentary, however, psychologist David C. Geary of the University of Missouri at Columbia argues that Evans' analysis is actually consistent with male/female differences observed in many species, as documented in a large body of research.
"Of course, driving is not in our evolutionary history, but in comparison to girls and women, a tendency for boys and men to approach life in a more aggressive and risky manner is a reflection of our evolutionary historymale-male competition in particular," Geary writes. "The sharp increase in the tendency to engage in aggressive, risk-taking behaviors in adolescence and lack of driving skill result in an unfortunate and often deadly combination, as Evans demonstrates."
Evans notes that, even if differences in traffic fatalities originate in "fundamental immutable biological differences," there's still a lot that can be done to lower the overall rate of fatal crashes. He points out that countries such as Great Britain, Canada, and Australia have achieved significant reductions in this rate over the last decade or so, whereas the United States has not.
"The aim is not the futile one of getting young men to behave like old ladies," Evans says, "but . . . to reduce risk to all road users."
Copyright © 2006 by Ivars Peterson
Evans, L. 2006. Innate sex differences supported by untypical traffic fatalities. Chance 19(No. 1):10-15.
______. 2004. Traffic Safety. Science Serving Society. See http://www.scienceservingsociety.com/traffic-safety.htm.
Geary, D.C. 2006. Commentary on "Innate sex differences supported by untypical traffic fatalities." Chance 19(No. 1):16-17.
Giesbrecht, L., J. Bose, and M.P. Cohen. 2006. Measuring exposure to vehicle crashes. Chance 19(No. 1):22-27.
Meyer, M. 2006. Commentary on "Innate sex differences supported by untypical traffic fatalities." Chance 19(No. 1):18-19.
Leonard Evans has a Web site at http://www.scienceservingsociety.com/.