Baltimore Spokes
Biking in Baltimore
Sign Up!
Welcome to Baltimore Spokes
Saturday, March 25 2017 @ 01:52 AM UTC

People in car crashes vs. people in bike crashes for Baltimore City by age.

Biking in BaltimoreThe question came up just how safe is it to bike in Baltimore? So I thought some graphs were in order:

Background: It has been observed the more people drive (Vehicle Miles Traveled or VMT) the more traffic crashes there were and conversely the lower VMT (as has been observed in recent years) the fewer traffic crashes. So this has lead to the notion of "exposure", that is to say the more there is of something out there the more accidents there will be. This works fine for automobiles but not so well for bike/ped issues.

There are some stats that show when VMT goes does down bike and pedestrian crashes go down as well and there are stats that show despite increasing the number of cyclists (increasing their exposure) the crash and fatality counts remain near constant.

Comments on the charts: So the way I look at things are the charts on the left, which is just comparing raw counts. I supplied the charts on the right for those who insist there is no such thing as the safety in numbers phenomena and insist that the more cyclists, the more crashes. So if there where the same number of cyclists as drivers at the current crash ratios per mode share were maintained, this is how it would look.

Of course what stands out is biking for kids in this city is dangerous (more K-12 bike education please), otherwise not really dangerous if you know the rules of the road. I'l agree there are some stinky driver's attitudes out there but they don't make the place unsafe, just annoying.

Again, I think the chart on the left represent relative risk of injury but even if we greatly exaggerate our small numbers to get them on par with the number of drivers, the chart on the right comes out near 50:50 with some wild fluctuation (due to the small sample size of cyclists.) Of course I am ignoring the issue for younger kids, which is more complex then what I can give it justice here.

Again, I think the chart on the left represent relative risk of death. The chart on the right is what one cycling death to 28 that died in cars in Baltimore looks like with the same "exposure". Over the years different aged cyclists have died so that age slot is not really representative of what to expect in the future but on average just one cyclist has died per year in Baltimore. Yes that is a tragedy but is it worst then the 28 that have died in a automobile?

So if I have failed to convince your relative risk is the charts on the left, the fatality chart on the right is the worse I can realistically contrive and even so there are a lot of empty age bands with no cyclists deaths but with people in cars deaths. Also bear in mind if we double the cycling mode share (with near the same crash counts) the red bars on the right charts would be cut in half. Baltimore is behind the curve for a major city bike lane miles as well as the bike mode share so I hope over time as we get more cyclists out on the streets even the charts on the right will look more favorable for cyclists.

In conclusion: If you are still feeling intimidated cycling in Baltimore, please see our collection of links Must read for bike safety

Crash data from: MHSO Select Benchmark Reports
Mode share from: American Community Survey, Commuting Characteristics by Sex

Story Options


Trackback URL for this entry:

No trackback comments for this entry.


The following comments are owned by whomever posted them. This site is not responsible for what they say.
Authored by: Anonymous User on Tuesday, March 13 2012 @ 10:45 AM UTC People in car crashes vs. people in bike crashes for Baltimore City by age.
I generally agree that risk is comparable or less by bike vs car and when we do so much more by car we put ourselves at more risk....but I am not sure that that is what these graphs show. I am glad you put this together...but perhaps it could be tweaked to better get at what you are trying to show.

These are not demographics broken down by individuals correlating individuals miles traveled by bike or by car and those individuals' crash/injury/fatality rates. These are "age groups" avg miles traveled by bike/car and "age groups" crashes/injuries/fatalities
these graphs help us understand individual risk only so far as we fit the average person of our age group.
If we also know that age group is a very very good predictor of bike/car crash/injury/fatality....then these graphs could reasonably extrapolate to individual risk....
but while we know that children vs adults age grouping is useful at predicting accident risk in general...beyond that I am not sure that breaking down crash data by age groups allows us meaningful predictions of risk.

What it does show is that amongst adult age groups relative car vs bike risk is very roughly comparable to that age group's relative car vs bike mileage (ie risk is 1:1 or 50:50)

If it turned out that the adult age groups with the highest bike mileage also had the highest bike/car crash ratios, then this data might actually suggest the opposite. Perhaps if we plot the crash/injury rate as a function of the average mode share of different age groups...then that might tell us age-group risk is really a surrogate for mode share risk. The absence of a meaningful correlation would tell us that either there is no mode-share risk correlation (ie more biking amongst an age group does not yield higher risk) or that age and mode share interact in opposite fashion to counter each other out (ie more biking does yield higher risk...but only within an age group, and age moderates this effect).

The point is...if we want to show that more bikes on the street will make biking safer, then introducing a third variable (age) only confuses things.
I say confuses because it complicates both statistics and communicating this to the average person.
I understand its hard to convince baltimore skeptics by pointing to other cities (with higher bike use and comparable crash rates)
and I also understand it is physically impossible to not have a third variable (or many more that just go un-measured or unaccounted).
But since age does play a role in risk, and everyone knows it does, I think this may confuse more than it helps.

Perhaps baltimore mode share and crash statistics as a function of time?

but most important is to show mileage on the x and crashes on the y...people understand that.
[ ]
Thanks for taking the time to comment, always appreciated.

Safety per exposure:
When we talk about relative "safety" per exposure we do not know if it is because the roads are safer, cars are safer or just people are driving more. The best I can figure out it's all of the above. So are people who live in Frederick and work in DC safer then those who live and work in DC simply because they drive more?

In contrast we have the Allstate report that shows the average frequency of car crashes and it shows Baltimore with an average close to one every 5 years while the national average is closer to one every ten years. This to me shows relative safety a lot more accurately then crash per VMT figures (which MD is close to the national average.) People who drive do basically the same stuff day to day and VMT really has nothing to do with day to day going to work and shopping. Similarly cyclists do basically the same stuff day to day, especially in the confines of the city. Granted cyclists may not be doing exactly the same thing as motorists but a figure that is close approximation to a crashes per motorist doing whatever they do vs. crashes per cyclist doing whatever they do is... well the best I can do with available data that's the same for both modes. But that is not to say I would not be interesting in seeing different comparisons with different data but we just don't have total bike miles like we do for motorists.

Why the age groupings:
The short answer is I wanted to separate kid issues from young adult issues from people with more experience. The sad truth is kids are over represented in Baltimore bike crashes but then again they are not doing so well in car crashes either. On the motoring side the 20 something age is over represented. How does this info influence your decision to bike or drive if you fall into those age groups? I guess that depends if you buy my argument for the graphs on the left but if not I tried to do my best in presenting relative risk based on exposure that is the same for both modes. But till someone takes the time to do a good study on this subject this is going to have to be the best hint we have.
[ ]

My Account

Sign up as a New User
Lost your password?


Site Map


There are no upcoming events

Older Stories

Sunday 12-Mar

Saturday 11-Mar

Friday 10-Mar

Monday 06-Mar

Sunday 05-Mar

Tuesday 28-Feb

Sunday 26-Feb