Replies to This Discussion

Permalink Reply by yaj 7.4 on March 5, 2014 at 3:32am

Interesting article - too bad only one US city discussed = NY.

The article states:

It can be difficult to compare safety between cities because of the lack of consistency in data collection and because of the need to frame injuries and deaths within the context of 'exposure' – the overall numbers of trips, total distance or time spent cycling. Under-reporting of cycling crashes is also a well-documented problem. Nevertheless, Pucher and Buehler's book listed figures for annual fatalities per 10,000 daily cyclists:

• Copenhagen 0.3
• Amsterdam 0.4
• Vancouver 0.9
• Toronto 1.3
• Portland 1.9
• Montréal 2.0
• Paris 8.2
• London 11.0
• New York 37.6

Can that be correct? Yikes!

• ▶ Reply

Permalink Reply by Lisa Curcio on March 5, 2014 at 3:47am

I just ordered the book because I want to see how they collected their data. :-)  I cannot imagine that there were 37.6 cyclists killed in 2009 for every 10,000 daily cyclists.

yaj 7.4 said:

Interesting article - too bad only one US city discussed = NY.

The article states:

It can be difficult to compare safety between cities because of the lack of consistency in data collection and because of the need to frame injuries and deaths within the context of 'exposure' – the overall numbers of trips, total distance or time spent cycling. Under-reporting of cycling crashes is also a well-documented problem. Nevertheless, Pucher and Buehler's book listed figures for annual fatalities per 10,000 daily cyclists:

• Copenhagen 0.3
• Amsterdam 0.4
• Vancouver 0.9
• Toronto 1.3
• Portland 1.9
• Montréal 2.0
• Paris 8.2
• London 11.0
• New York 37.6

Can that be correct? Yikes!

• ▶ Reply

Permalink Reply by Lisa Curcio on March 8, 2014 at 2:39pm

I just got the book in the mail.  First of all, for New York, Tokyo, Paris and London, it is fatalities per 100,000 cyclists, not 10,000.  Also, here is the description of how they come up with the numbers:

The fatality rate shown here is calculated as average annual fatalities over a five-year period divided by population weighted by bike mode share.  Thus, it is only a rough approximation of the fatality rate relative to different exposure levels in the four cities.

So the numbers mean something relatively, but not absolutely.

The numbers for the other cities is per 10,000 cyclists, and there is a number for Chicago in the book--4.2.

However, again the numbers are not exactly cast in stone:

Because the number of cyclist fatalities fluctuates from year to year, we calculated the average number of fatalities over the last five years for each city.  For the exposure rate, we used the number of daily commuter cyclists because those data are derived from large census surveys that can be disaggregated to the city level.  There is no other source of internationally comparable and statistically reliable date on cycling levels in each city.  The problem with this methodology is that the number of fatalities in the numerator is due to cycling for all trip purposes, whereas the number of cyclists in the denominator includes only work commuters. 

The emphasis on the last sentence is mine.  The other problem with the data is that the census forces the person answering to designate a "primary" mode of transportation so someone who bikes and takes transit but decides his/her primary mode is transit will not be identified as a "work commuter".

Again, the numbers might be useful in a relative sense, but not an absolute sense.

• ▶ Reply

Permalink Reply by mike w. on March 8, 2014 at 4:18pm

Statistics are like bikinis: What they reveal is interesting, what they conceal is essential.

• ▶ Reply

Permalink Reply by Mike Zumwalt on March 8, 2014 at 4:25pm

Why was the first pic of an apparent "fixie" doing a blurred speed? 

Studies have also shown that road rage and a sedentary lifestyle can cause you to stress out and die early.

Maybe we should all drive to the gym and ride a nice stationary bike while watching CNN?

• ▶ Reply