The next two columns of Table 1-1 s

The next two columns of Table 1-1 show the count of all fatalities to these occupants and the percentage of all fatalities that were frontal. Overall fatalities decreased from 30,336 to 25,663 – i.e., by a lesser proportion than frontal fatalities. As a consequence, the proportion of fatalities that are frontal decreased from 51.4 percent to 45.4 percent. While this is in the right direction, it is difficult to draw conclusions from the trend line. There have been safety improvements addressing side impacts (dynamic test, side air bags) and rollovers (belt use) as well as frontals, so there is no particular reason that frontal fatalities should change as a proportion of all fatalities. Furthermore, factors other than intentional safety improvement can influence the trend. For example, older drivers and smaller cars tend to increase side-impact fatalities, and fortuitously reduce the proportion of fatalities that are frontal. The shift from cars to SUVs tends to increase rollovers, and again reduce the proportion of fatalities that are frontal.