A word of caution with these statistics and maps: you need to consider crashes per capita in each of those areas, rather than the total number. In areas with more people, and therefore more roadusers, you expect (unfortunately) to have a higher number of crashes.
Consider a city with a population of 10,000,000 and a murder rate of 100 people per year compared to a town with a population of 100 and a murder rate of 90 people per year. Is the town safer because the total number of people murdered is lower?
If anyone has the ability to produce a map of crashes/fatalities per roaduser/ped/cyclists, I'd be very interested to see the results.
Rach Stevo07.7.11 - 11:24 am
reply
Consider a city with a population of 10,000,000 and a murder rate of 100 people per year compared to a town with a population of 100 and a murder rate of 90 people per year. Is the town safer because the total number of people murdered is lower?
Yeah, I think so, given the town consists of ten people, I'm willing to bet it's relatively safe.
markd responding to a
comment by Rach Stevo
07.7.11 - 3:57 pm
reply
mutually assured destruction would make it safe at that point....
Roadblock responding to a
comment by theroyalacademy
07.7.11 - 7:48 pm
reply
Holy crap, 4 people died on my street. In Chatsworth.
I wish this map differentiated between cyclists and pedestrians. Not that either would be acceptable, but I'm mostly concerned for my safety as a cyclist rather than pedestrian.
outerspace07.7.11 - 9:37 pm
reply
Or one murderer who's been so busy that he hasn't got around to the final nine victims.
But seriously - great article, Josef!
mr rollers responding to a
comment by theroyalacademy
07.7.11 - 10:12 pm
reply
It would be nice but realize that streets that are safe for pedestrians are safe for cyclists.
Roadblock responding to a
comment by outerspace
07.7.11 - 11:17 pm
reply
I think the total number of deaths is a much more appropriate measure to use here than deaths per capita.
Sure, if you, as an individual cyclist, want to know which streets are the most dangerous to ride on, then measuring cyclist fatalities per capita will give you some idea* of the risk involved.
But if you are a planner trying to decide how best to reduce cyclist fatalities, I think you'd want to know the number of fatalities
per area, because your most important tools are spatially constrained: miles of bike lanes, widths of sidewalks, etc.. With limited resources, you'll want to focus your efforts on areas where you can save the greatest number of lives.
Yes, the pattern of fatalities is partly driven by population density, but where there are more riders, there should be better infrastructure. The point that the LACBC is making with their maps is that the areas with the highest population densities and highest numbers of fatalities actually have the
fewest bike lanes. And the fact that these areas are also the poorest means that many of the people living there ride a bike because they have no other choice.
*A much better measurement would be the number of fatalities per cyclist-mile ridden, broken down at the street level, but reliable data would probably be difficult to come by.
nathansnider responding to a
comment by Rach Stevo
07.8.11 - 5:56 pm
reply
I agree that money, infrastructure, and planning efforts should be aimed at those areas with the highest total number of collisions.
My concern is drawing the conclusion that the risk is higher in, for example, a low-income neighborhood. It may well be, but in order to properly assess that risk you need to divide by the number of cyclists (or miles ridden, as suggested).
From the LACBC post:
"crash fatalities are unevenly distributed in areas of higher density and lower-income neighborhoods"
I believe to justify this statement you need to use crashes per capita, to compensate for the fact that there may be 10 times as many cyclists and drivers in dense areas compared to low-density areas. But funding should absolutely be directed towards those areas with the highest total number.
Rach Stevo responding to a
comment by nathansnider
07.9.11 - 3:26 pm
reply
-->
Ridazz Roulette!
