I posted earlier about flaws in IIHS fatality data that they admit; namely that data is not normalized for all demographics. About 60-70% of deaths are caused by male drivers, most of them young, probably drinking/speeding,etc.
If we banned all Hondas (great idea) for males under 30 and forced them into Expeditions or 750iLs would total death rates go down? Vast majority of big vehicles driven by females(minvans), they are richer and older than average drivers (>35). So is it the car or the driver? How but both.