I found this interesting study that shows the risk to drivers of other
vehicles vs the risk to drivers for different 1995-1999 vehicle
models.

http://www.lbl.gov/Science-Articles/...ty-newWin.html

For cars, it shows Camry to be the safest (with Accord and others
pretty close). The data is not normalized per mile traveled though.

What I find odd is that Prizm is considerably less safe than Corolla,
according to them. Is there a likely mechanical explanation (dual
airbags are standard in both, but perhaps the quality is different),
or is this a statistical artifact due to the poorer and thus younger
people buying Prizms?

By the way, does anyone know of a similar, but more up-to-date study?
I'd also like the probabilities of disablement included with the data
given per mile traveled.