Bottom line: Per mile, risk of death on a bicycle is about thirteen times higher than risk of death in a car/SUV/van. Calculation shown below.
Background
I read a story in the Washington Post today, about a woman who was killed while bicycling, in the bike lane, alongside River Road in Bethesda. Crushed by a careless commercial truck driver making a right turn. Leaving behind a husband and two young kids.
The truck driver was given the maximum sentence allowed by law, which in this case was a $2000 fine. And a brief moment of shame in court. He’s still on the road, driving a truck.
I looked up the accident scene on Google, and it was the worst kind of grudging, cheap, zero-effort retrofit urban bike lane that the very least of your tax dollars can produce. Based on historical images, they took a narrow, disused shoulder of a 35-MPH urban arterial highway, and painted little arrows and bike icons in it.
Well, there’s your bike lane, right there. Problem solved. The result of that zero-effort accommodation of bicyclists is every bit as safe as you might reasonably expect.
Above, the middle red circle is the site of death. It’s just one of many driveways opening into the busy commercial establishments that line the road. It’s located just 500 feet from the Capital Crescent Trail, a dedicated bike path whose road-crossing bridge can be seen circled in the background. Which is almost certainly why they bothered to re-paint the road shoulder.
In the foreground is a sign. Based on Google street view, that sign was only placed there a few months ago. If you don’t routinely bike in an urban area, you’d think that sign was there to remind motorists to use caution, and look before they turn. But for that purpose, in this context, a sign like that is useless. Motorists don’t even perceive signs like that, in the crowded visual field of an urban motorway, at 35 MPH and up. The actual, practical purpose of the sign is to warn bicyclists and pedestrians that they are in the middle of a war zone, and that they should act accordingly if and as possible.
But in this situation, there’s not much a bicyclist can do. The bike lane appears to be just under 3′ wide. The curb is about a foot to your right, so there’s no escape in that direction. Your life depends on the caution and good sense of the drivers passing a couple of feet to your left.
A slender reed, for sure.
But is it really as dangerous as it looks?
Yes it is.
And, for some reason, this appears to be an answer that absolutely infuriates bicycle advocates. Not because they don’t want to make roadways safer for bikes. But simply because they don’t want to believe that there are significant downsides to bicycle transport in America. This, even if everybody grasps what a drag it is to (e.g.) bike in bad weather.
(And, to be clear, I’m a lifetime bicycling enthusiast. But I’m also a realist.)
Years ago, I did the homework to answer this question, to my own satisfaction. I came up with an estimate that bicycling is about ten times as dangerous as driving, per mile traveled. That’s in terms of risk of death. It’s even higher in terms of risk of injury requiring medical attention. Those specific calculations are lost in the mists of time. So I thought I might update and document that here.
Any estimate of bicycling safety ends up combining data from two separate sources. That’s always a risk for accuracy, but it is what it is.
Information on traffic-related deaths comes from a motor vehicle crash reporting system maintained by the National Highway Transportation Safety Administration, the Fatality Analysis Reporting System. Whatever its limitations, that’s the U.S. gold standard for counting traffic-related fatalities in the U.S.
(Secondarily, if you have an interest, you can use nationally-available hospital statistics from the U.S. Public Health Service’s Healthcare Cost and Utilization Project to find the number of hospitalizations and outpatient visits related to bicycle accidents. But non-fatal injuries actually show an even grimmer picture for bicycling versus driving, so it’s probably sufficient to settle on an estimate of risk-of-death while bicycling versus driving, per mile.)
So, bicyclists accounted for about 2.4% of all U.S. traffic deaths in 2020. That figure is roughly constant over time.
Information on bicycle-miles traveled comes from the National Household Travel Survey (NHTS). This is based on a log-diary survey of thousands of U.S. households, capturing how and why they traveled over the course of a week. Near as I can tell, it’s the only nationally-representative information on actual use of bicycles versus other modes of transportation.
Source: 2017 National Household Travel Survey
Based on that large, nationally-representative log-diary survey, bicycle transport accounted for 0.2% of all U.S. household transportation, or about 8.5 billion bicycle-miles per year in 2017. (Sounds like a lot until you realize there are about 330 million U.S. residents, so that works out to about 25 bike-miles per person per year.)
We have to re-calculate the percentages above, to restrict this to cars versus bikes. After that, it’s just simple math.
Summary.
Yes, bicycle transportation really is as dangerous as it looks, as you drive along the road.
Using two gold-standard U.S. databases, bicycling appears vastly riskier than driving a car. In this most recent calculation, I estimate about 13 times higher risk of death, per mile, on a bike, versus in a car. That’s reasonably consistent with the estimate I got years ago.
And, as I recall, if you expand to non-fatal injuries requiring medical attention, the relative risk is actually much higher. More like 50 or 60 times the risk, per mile. That’s for the obvious reason that a collision that produces only minor bumps and scrapes to car occupants can produce severe wounds to an unprotected bicyclist.
Four things are worth noting.
First, the absolute risk is low. If you bicycle 1000 miles a year — which is a lot — your risk of death-via-bike is about 0.01% per year.
Second, for older adults, the exercise benefit vastly outweighs the crash risk. This is another one that I did the homework years ago, then lost the analysis somewhere. For the average 65-year-old man, all-causes risk of death is about 2% per year. Best available research suggests that frequent vigorous exercise roughly cuts that in half. The health benefits of frequent bicycling likely outweigh the risk-of-traffic-death by an order of magnitude or two.
Third, on paper at least, walking around traffic is more hazardous than bicycling. I’m not sure to the extent this is driven solely by work-related pedestrian accidents in big cities. But whatever the cause, this too seems plausible. In effect, we let amateurs drive 3-ton pieces of machinery, at high speed, around crowds, with virtually no enforcement of rules or penalties for engaging in risky behavior. It’s a wonder that so few pedestrian deaths result.
Finally, if you do a deep dive in the FARS database, you’ll find that dead pedestrians and dead bicyclists have something in common with dead drivers. An astounding fraction of them are dead drunk at the time. Roughly half, in each case. The moral of the story is that BUI and WUI are maybe not as deadly to others as DUI, but they clearly up your risk of death on the roadways.
But in this case, that’s irrelevant. It was broad daylight, and the victim was returning from a school function for her kids. The death was just the result of a toxic combination of thoughtless, zero-effort bike lane design, and bad luck. Ten seconds sooner, or ten seconds later, and she’d have been fine. It’s just an unavoidable risk of bicycling in most urban areas of the U.S.
