The Oregonian reports  on Portland having received a rating as one of the safest walking cities in America—Portland was #9. But is it really a safe walking city?
The methodology is seriously flawed. It is clearly designed to make a political point and not provide meaningful data. The way they compiled the ranking was to divide the number of pedestrian deaths divided by the number of people who walk to work. What you would ideally want is some measure of total pedestrians or, better yet, total pedestrian-hours.
Proxying this by using walking to work may seem like a good idea, but I doubt that it is. Clearly walking to work is very, very highly correlated with urban density but may not be correlated with the population who walk for any reason at any time. So you skew the statistics and, not surprisingly, very dense cities look safe and sprawling cities do not. Which is, I imagine, their point.
Now, I hate sprawl as much as the next guy, but it is not necessarily bad for walking safety in general. Yes, you can’t walk to get everywhere, but walking around inside the little cloistered neighborhoods that were the rage before the new urbanism ethos kicked in is probably pretty safe.
What sprawl is bad for is walking in general. I don’t think conflating the two is helpful.