Injuries from crashes statistics

Unlike the rate per miles driven, in the U.S. fatalities per population has stayed fairly constant with only a 5% drop from 1923 to 2000. Why the great disparity in the behavior of the two statistics One possibility is that most of the improvement in the rate per miles driven is due to increase in travel rather than to a reduction in number of crashes. Thus a road segment may be equally safe (or unsafe) regardless of the number of cars traveling on it (within limits) and a car may be equally safe (or unsafe) regardless of the miles driven. Another possibility, raised by Sivak (2002) is that a society has a certain tolerance to fraffic injuries, not in absolute terms (because the absolute numbers keep increasing) but relative to population size. 

Multiple independent analyses of seat belt effectiveness, using crash data from different countries, and different statistical methods have all yielded quite similar results, reducing fatalities and injuries by approximately 40-50 percent, as detailed in Table 10-1 (WHO, 2004). However, the exact numbers in the table, should be considered as potentially inflated estimates because they are based on police reports, and these are known to inflate the percent of belt use. This is because drivers - when they can get away with it - will report to the officer that they had their belt on (Li et al., 1999). Consequently as the injury severity decreases, the amoimt of over-estimation of belt use is most likely to increase. 

---