Comment by tonmoy

While their conclusion is probably correct, I would have liked to seen the number of fatalities normalized by population, miles driven, number of pedestrian increase, speed limit change etc