JonnyHK 

Risky cycling rarely to blame for bike accidents, study finds

Cyclists disobeying stop signal or wearing dark clothing at night rarely cited in collisions causing serious injury

Peter Walker
guardian.co.uk, Tuesday 15 December 2009 16.25 GMT

A tiny proportion of accidents involving cyclists are caused by riders jumping red lights or stop signs, or failing to wear high-visibility clothing and use lights, a government-commissioned study has discovered.

The findings appear to contradict a spate of recent reports speculating that risky behaviour by riders, such as listening to music players while cycling, could be behind a near 20% rise in cyclist deaths and serious injuries in the second quarter of this year.

The study, carried out for the Department for Transport, found that in 2% of cases where cyclists were seriously injured in collisions with other road users police said that the rider disobeying a stop sign or traffic light was a likely contributing factor. Wearing dark clothing at night was seen as a potential cause in about 2.5% of cases, and failure to use lights was mentioned 2% of the time.

The figures were slightly higher when the cyclist was killed, but in such cases only the driver's account is available.

The data, which was analysed by the Transport Research Laboratory (TRL), showed that more than a quarter of all cycling deaths in 2005-07 happened when a vehicle ran into the rear of a bike. This rose to more than one-third in rural areas and to 40% in collisions that took place away from junctions.

The 64-page analysis found that police attributed responsibility for collisions more or less evenly between drivers and cyclists overall, but this was skewed by the fact that when child riders were involved their behaviour was named as a primary factor more than three-quarters of the time.

With adult cyclists, police found the driver solely responsible in about 60%-75% of all cases, and riders solely at fault 17%-25% of the time.

The cyclists' lobby group CTC said the report showed that the government needed to focus more on driver behaviour rather than on issues such as cyclists wearing helmets. The TRL published a separate DfT-commissioned report today in which it was estimated that the universal use of helmets could save between 10 and 15 lives a year, a conclusion disputed by the CTC.

"The main cause of crashes seems to be 'failed to look properly', whereas very few cyclists are injured or killed acting illegally, such as failing to use lights at night or disobeying traffic signals," said Chris Peck, from the lobby group.

"We believe this report strongly supports our view that the biggest problem for cyclists is bad driving. With that in mind we are greatly concerned that the government still seems fascinated with analysing and promoting cycle helmets, the value of which appears to be inconclusive. We believe that the government should now focus on tackling the causes of injury which appears to be mainly inconsiderate and dangerous driving. Reduced speed limits, stronger traffic law enforcement and cycle-friendly road design are the solutions."

TRL recommended that more research be carried out into the relatively high numbers of young casualties, finding that those aged 10 to 15 were most at risk of injury for each mile cycled. Riders aged 16 to 29 were more likely to suffer harm than any other adult group.

The data – which covered incidents on the highways – showed that 3% of all collisions leading to deaths or serious injuries took place on bike lanes, and almost 80% of casualties happened during daylight hours. Just over 15% of all such accidents involved the cyclist alone.

BarracksSi

In short, then, driving -- or riding -- like an idiot can get you hurt.

Friggin' rookies, encouraged into cycling by $4/gal gas but with just as little sensibility as they had when they thought buying an SUV was a good idea.

I-Like-To-Bike

Any reference, besides your crystal ball, for claims that "Friggin' rookies" were previous SUV owners, were involved in the observed risky behavior, or that the price of gasoline ever affected the overall sensibility of cyclists, anywhere, anytime.

Please don't confuse your straw men fabrications with measured data, or even with anecdotes.

John E

On the surface, it seems scary that the motorist was at fault in most motor vehicle - bicycle mishaps, but this does not imply that we are always sitting ducks. It does place a premium on defensive cycling techniques.

damnable

+1 on focusing more on motorist behavior.

ghettocruiser

These numbers seem strangely at odds with certain agenda-pushers both here on BF and in local editorial boards.

But neither of these groups ever let the facts get in the way.

genec

Yeah some of those agenda pushers continue to fall back on some report written in the '70s, based on police reports in Santa Barbara, and evaluated by their guru as "gospel." Never mind any other report coming from any other area since. Any other report is viewed as somehow incomplete or biased... but never mind the bias and ancient data of the '70s report. (never mind the guy behind the curtain...)

Hell, cell phones did not even exist in the '70s... oh the times they have a'changed.

chipcom 

This is the only thing in the article that really jumped out at me as worthy of discussion.

genec

Yeah, sure seems to deny the old "overtaking collisions are rare" claim.

And other agencies report that rear end collisions make up the majority of auto accidents (NHTSA) (sorry, no link)

mikeybikes

I tried searching, but I really can't find the statistics (I'm probably doing it wrong).

If that's the case, that really changes things. Then the question comes: Are bike lanes really as bad as a lot of the VCers would like you to think?

chipcom 

Bike lanes are not bad things. People who expect painted lines to automagically protect them from harm or do not understand the potential pitfalls of bad implementations of bike lanes can be. To your point, sure, one might be able to speculate that additional space on the roadway, clearly marked for bicycle use only, might have helped to prevent at least some of these rear end incidents, but in no way could be considered some magic bullet that would have prevented all or even most of them.

I-Like-To-Bike

Everything relating to cycling risk provided by "a lot of the VCers" should be taken with a healthy dose of skepticism, if not out right scorn, if the claims/advice rely on Forester brand conclusions that are wholly based on his notoriously biased, self serving Forester Brand "tests", statistical juggling/misrepresentation of others' work, and sophomoric analysis of same.

Road Fan

This study

Basav Sen, Brittany N. Campbell, John D. Smith, Wassim G.Najm, "Analysis of Light Vehicle Crashes and Pre-Crash Scenarios Based on the 2000 General Estimates System", Performed by John A.Volpe National Transportation System Center, Cambridge, MA, Sponsored by National Highway Traffic Safety Administration, Washington D.C, November 2002, DOT HS 809 573

contains a pretty broad and comprehensive summary of collision frequency of occurrence broken down by collision type, and it shows that (Table 2-2(a)) for light vehicles, 29.4% of all crashes invoving light vehicles are rear-end. Next in line are called Crossing Path at 25.9%; this is an aggregate of a wide range of ways to get T-boned or to T-bone at formal or non-formal intersections. Third in line is Off-Roadway with 20.9%.

What is surprising is that those are not the biggest contributors to fatality. I can't find the US DoT studies right now breaking out fatalities by the same breakdown of crash types, but this recent early report


2008 Traffic Safety Facts Annual FARS/GES Report (EARLY EDITION)
The 2008 Traffic Safety Facts FARS/GES Annual Report (Early Edition) is now available and provides the most current 2008 data and final 2007 data. The “Early Edition” does not contain exposure data (i.e., vehicle miles traveled, registered vehicles, licensed drivers) and other data points that customarily are not available until later. This "Final Edition" will be available soon. Click Here to read the entire NCSA FARS/GES 2008 Early Edition (DOT-HS-811-170)

shows (Figure 22, chapter 4) the fatality rate for rear-end is about 2 per 1000 crashes and the injury rate is about 380 per 1000 crashes. The fatality rate is highest for Head-on, at about 29 per 1000.

Discussions I've had with the experts who analyzed this data and in my former activity in the ISO committee on collision avoidance systems indicate that the closing velocities in car to car rear end collisions are relatively low, in the 30 mph or less range, even at highway speeds. I can't find the NHTSA studies on this, might have been taken off the site for now.

Car to bike is of course much more severe because of the lack of protection the bike gives to the rider, compared even to a motorcycle, much less a car.

Road Fan

There are several sources for stats: www.nhtsa.gov and the Insurance Institute for Highway Safety are perhaps the most prominent. NHTSA has direct access to the federal compilation of police reports from all 50 stats. IIHS is able to get the same access, but tends to look at the data differently. IIHS also has access to insurance company reports on crashes, injury (or worse) and payouts.

genec

Right and remember we are always discussing fatalities. I myself have been involved in 3 car-bike collisions in which motorists were at fault, but I am not dead. So numbers of fatalities do not always convey the whole picture.

Neither do police reports... of the three collisions that I was involved in, only one had a police report. The other two were admitted by the driver and taken care of by insurance or out of pocket. No police reports were filed. (I only went to hospital for the one reported collision).

My point being that there are likely far more car-bike collisions that occur that are unreported and thus fail to make any study. Further, there are likely more cyclists injured by car-bike collisions than get reported. The propensity is for cyclists to pick themselves up, brush off, take account and limp on their way... that stuff never makes it into any report and therefore is not part of any study, regardless of fault.

genec

The thing I find interesting is that collision rates have gone up... but overall death rates are lower. Cars are safer now... but motorists keep crashing into each other and other things.

TRaffic Jammer

That's because the idiot behind the wheel is still in charge, whereas the safety systems are automatic. Wooohooo, I'm safer, now I can driver faster and look less.

After 20+ years of "risky" cycling I am a living testament as to how it can make the cyclist safer. I've often contended that I simply ride where cars aren't. This includes , down the middle of the urban core yellow line, jumping green lights, and running reds (just not so motorists have to bother with me). The almost offical take on here from the police is if you sorta stop and look and proceed safely they won;t burn you for the red light infraction. This is nothing new I cant believe it's taken them this long to figure out via this study. Talk about tinted with motorist coloured glasses. I've been saying that the "rules of the road" are what are getting us killed out there, because they are written for motorists.

JoeyBike

I'll add to that with 30 years of so-called risky cycling under my wheels.

The times I feel most at risk form motorists is, like the study indicates, when a vehicle is roaring up behind me. I am totally at their mercy and helpless to their lack of attention/ability/awareness. Except, of course to position and clothe myself to BE SEEN. I ride like I do to limit, or hopefully eliminated, cars overtaking me.

My other vulnerability is cross streets with stop signs when I am biking on a major arterial. For a split second, every crossing auto has a shot at me. I call that "hit man style" driving. They stop at the stop sign, look both ways, wait......wait....., then soon as I get to the point of no return - GO! Again, the cure is to BE SEEN by their brains, not their eyes. This involves lane positioning again - riding in the LEFT tire track of the right lane. The more red lights I run, and the faster I ride, the more time I can spend in that left tire track before drawing the ire of overtaking vehicles.

I have said this stuff a million times here. I have an entire blog devoted to this (going on 3 years now). As TRaffic Jammer wrote: "This is nothing new". They just were not asking the right questions to the "wrong" people.

crhilton

It wouldn't surprise me if auto accidents had lots of rear enders and bicycles didn't. Simply because following behind someone is what two cars do most of the time. With bicycles you're general overtaking as soon as possible: So there's little time to sit in and relax while trying to match his speed.

I am surprised that this report indicates that rear enders are common. It strikes me as odd not just because I tend to believe rear enders are 2-5%, a la John (I hate cyclists) Forester, but because I watch cycling accident deaths in the news and I rarely see a rear ender. Obviously this isn't good statistical evidence, even for me, but that leads me to be suspicious of their results.

I'm gonna keep an eye on this one. It betrays "facts" that I knew.

I-Like-To-Bike

Relatively few cyclists are involved in rear enders, because relatively few cyclists routinely take lanes in busy traffic and expose themselves to the risk of being rear ended, especially on higher speed roads. If any of that small subset of cyclists does get involved in a rear ender the injuries are frequently catastrophic. Most, if not all, Forester Brand analysis of cycling risk ignores the exposure, probability, and severity components of various collision scenarioes, and just counts "crash" totals in order to reach his desired conclusion.

crhilton

Am I misreading the link. It seems to say that more than 25% of serious injury accidents for cyclists are rear enders. That goes against what I thought, and what you just said. Although your criticism of Forester's analysis of other statistics does seem like a plausible criticism.

I-Like-To-Bike

Most accidents do not result in serious injuries, in many, no injuries occur at all. That doesn't hold true for the rear end type collisions.

Certainly that 25% number sounds right for a population of cyclists where lane takers are probably in the low single digit % slide of the population. I'm sure that over representation would be found in the fatality rates too where rear end collisions are involved ina sizable proportion of deaths, even though a relative small proportion of cyclists are exposed.

Road Fan

The reports I cited were only about collisions involving motor vehicles and NOT bicycles or pedestrians. Rear-end crashes were those in which a MV hit another MV. While it is possible there's a lot more car/bike events than are reported, I don't think non-reports are that likely in the case of MV/MV.

Besides, in studying auto safety or road safety based on collision data, you are stuck looking where the data is. Non-reported crashes are NOT REPORTED, by definition. There's no way those events can even be proved to have occurred. Reports have to be viewed (by an investigator) as anecdotal.

Besides #2: If you are the only one telling me (the investigator) that you had three crashes that were not reported, how can I look at that as changing anything, when I have data representing 6 million plus/minus annual collisions available, and over 10k case studies available in FARS?

This doesn't say that your info is wrong, just that investigations have to look where the data is.

Road Fan

For car/car rear end collisions, most do not result in injury. I'm not sure if this is what you're talking about, but in this thread we've been talking about car/car and car/bike.

meanwhile

You shouldn't assume that UK research will translate directly to the US. Or that this research is more valid than the research that it contradicts.