[marciero]

These very fatality numbers were discussed at length in a recent thread as I recall. Using raw fatality numbers to draw conclusions can be very misleading. The "safety in numbers" graph in the article-there is way too much variation to draw any conclusions. Even visual inspection of the fatalities graph seems to indicate decreasing or flat fatality rates with significant variability. That said, the numbers for individual cities is of interest. The actual report (I actually downloaded it) seems to have a lot of this kind of thing which is of interest, but nothing to draw sweeping overall conclusions with.